Qatar Foundation Annual Research Conference Proceedings Volume 2016 Issue 1

Data analytics is at the core of any organization that wants to obtain measurable value from its growing data assets. Data analytic tasks may range from simple to extremely complex pipelines, such as data extraction, transformation and loading, online analytical processing, graph processing, and machine learning (ML). Following the dictum “one size does not fit all”, academia and industry have embarked on a race of developing data processing platforms for supporting all of these different tasks, e.g., DBMSs and MapReduce-like systems. Semantic completeness, high performance and scalability are key objectives of such platforms. While there have been major achievements in these objectives, users are still faced with many road-blocks.

MOTIVATING EXAMPLE

The first roadblock is that applications are tied to a single processing platform, making the migration of an application to new and more efficient platforms a difficult and costly task. As a result, the common practice is to re-implement an application on top of a new processing platform; e.g., Spark SQL and MLlib are the Spark counterparts of Hive and Mahout. The second roadblock is that complex analytic tasks usually require the combined use of different processing platforms where users will have to manually combine the results to draw a conclusion.

Consider, for example, the Oil & Gas industry and the need to produce reports by using SQL or some statistical method to analyze the data. A single oil company can produce more than 1.5TB of diverse data per day. Such data may be structured or unstructured and come from heterogeneous sources, such as sensors, GPS devices, and other measuring instruments. For instance, during the exploration phase, data has to be acquired, integrated, and analyzed in order to predict if a reservoir would be profitable. Tens of thousands downhole sensors in exploratory wells produce real-time seismic structured data for monitoring resources and environmental conditions. Users integrate these data with the physical properties of the rocks to visualize volume and surface renderings. From these visualizations, geologists and geophysicists formulate hypotheses and verify them with ML models, such as regression and classification. Training of the models is performed with historical drilling and production data, but oftentimes users also have to go over unstructured data, such as notes exchanged by emails or text from drilling reports filed in a cabinet. Therefore, an application supporting such a complex analytic pipeline should access several sources for historical data (relational, but also text and semi-structured), remove the noise from the streaming data coming from the sensors, and run both traditional (such as SQL) and statistical analytics (such as ML algorithms).

RESEARCH CHALLENGES

Similar examples can be drawn from other domains such as healthcare: e.g., IBM reported that North York hospital needs to process 50 diverse datasets, which are on a dozen different internal systems. These applications show the need for complex analytics coupled with a diversity of processing platforms, which raises several challenges. These challenges relate to the choices users are faced with on where to process their data, each choice with possibly orders of magnitude differences in terms of performance. For example, one may aggregate large datasets with traditional queries on top of a relational database such as PostgreSQL, but the subsequent analytic tasks might be much faster if executed on Spark. However, users have to be intimate with the intricacies of the processing platform to achieve high efficiency and scalability. Moreover, once a decision is taken, users may still end up tied up to a particular platform. As a result, migrating the data analytics stack to a different, more efficient processing platform often becomes a nightmare. In the above example, one has to re-implement the myriad of PostgreSQL-based applications on top of Spark.

RHEEM VISION

To tackle these challenges, we are building RHEEM, a system that provides both platform independence and interoperability across multiple platforms. RHEEM acts as a proxy between user applications and existing data processing platforms. It is fully based on user-defined functions (UDFs) to provide adaptability as well as extensibility. The major advantages of RHEEM are its ability to free applications and users from being tied to a single data processing platform (platform-independence) and provides interoperability across multiple platforms (multi-platform execution).

RHEEM exposes a three-layer data processing abstraction that sits between user applications and data processing platforms (e.g., Hadoop or Spark). The application layer models all application-specific logic; the core layer provides the intermediate representation between applications and processing platforms; and the platform layer embraces all processing platforms. In contrast to DBMSs, RHEEM decouples physical and execution levels. This separation allows applications to express physical plans in terms of algorithmic needs only, without being tight to a particular processing platform. The communication among these levels is enabled by operators defined as UDFs. Providing platform-independence is the first step towards realizing multi-platform task execution. RHEEM can receive a complex analytic task, seamlessly divide it into subtasks and choose the best platform on which each subtask should be executed.

RHEEM ARCHITECTURE

The three layers separation allows applications to express a physical plan in terms of algorithmic needs only, without being tied to a particular processing platform. We detail these levels below.

Application Layer. A logical operator is an abstract UDF that acts as an application-specific unit of data processing. In other words, one can see a logical operator as a template whereby users provide the logic of their analytic tasks. Such abstraction enables both (i) ease-of-use by hiding all the implementation details from users, and (ii) high performance by allowing several optimizations, e.g., seamless distributed execution. A logical operator works on data quanta, which are the smallest units of data elements from the input datasets. For instance, a data quantum represents a tuple in the input dataset or a row in a matrix. This fine-grained data model allows RHEEM to apply a logical operator in a highly parallel fashion and thus achieve better scalability and performance.

Example 1: Consider a developer who wants to offer end users logical operators to implement various machine learning algorithms. The developer defines five such operators: (i) Transform, for normalizing input datasets, (ii) Stage, for initializing algorithm-specific parameters, e.g., initial cluster centroids, (iii) Compute, for computations required by the ML algorithm, e.g., finding the nearest centroid of a point, (iv) Update, for setting global values of an algorithm, e.g., centroids, for the next iteration, and (v) Loop, for specifying the stopping condition. Users implement algorithms such as SVM, K-means, and linear/logistic regression, using these operators.

The application optimizer translates logical operators into physical operators that will form the physical plan at the core layer.

Core Layer. This layer exposes a pool of physical operators, each representing an algorithmic decision for executing an analytic task. A physical operator is a platform-independent implementation of a logical operator. These operators are available to the developer to deploy a new application on top of RHEEM. Developers can still define new operators as needed.

Example 2: In the above ML example, the application optimizer maps Transform to a Map physical operator and Compute to a GroupBy physical operator. RHEEM provides two different implementations for GroupBy: the SortGroupBy (sort-based) and HashGroupBy (hash-based) operators from which the optimizer of the core level will have to choose.

Once an application has produced a physical plan for a given task, RHEEM divides the physical plan of a task into task atoms, i.e., sub-tasks, which are the units of execution. A task atom (a part of the execution plan) is a sub-task to be executed on a single data processing platform. It then translates the tasks atoms into an execution plan by optimizing each task atom according to a target platform. Finally, it schedules each task atom to be executed on its corresponding processing platform. Therefore, in contrast to DBMSs, RHEEM produces execution plans to run in multiple data processing platforms.

Platform Layer. At this layer, execution operators define how a task is executed on the underlying processing platform. An execution operator is the platform-dependent implementation of a physical operator. RHEEM relies on existing data processing platforms to run input tasks. In contrast to a logical operator, an execution operator works on multiple data quanta rather than a single one. This enables the processing of multiple data quanta with a single function call, hence reducing overhead.

Example 3: Again in the above ML example, the MapPartitions and ReduceByKey execution operators for Spark are one way to perform Transform and Compute.

Defining mappings between execution and physical operators is the developers’ responsibility whenever a new platform is plugged to the core. In the current prototype of RHEEM, the mappings are hard-coded. Our goal is to rely on a mapping structure to model the correspondences between operators together with context information. Such context is needed for the effective and efficient execution of each operator. For instance, the Compute logical operator maps to two different physical operators (SortGroupBy and HashGroupBy). In this case, a developer could use the context to provide hints to the optimizer for choosing the right physical operator at runtime. Developers will provide only a declarative specification of such mappings; the system will use them to translate physical operators to execution operators. A simple and extensible operator mapping is crucial as it enables developers to easily provide extensions and optimizations via new operators.

PRELIMINARY RESULTS

We have implemented two applications on top of RHEEM, one for data cleaning, BigDansing [1], and one for machine learning. The performance of both applications are encouraging and already demonstrate the advantages of our vision.

Our results show that, in both cases, RHEEM enable orders of magnitude better performance than baseline system. These improvements come from a series of optimization done at the application layer as well as at the core layer. As an example of optimization at the core layer, we extended the set of physical operators with a new physical operator for joins, called IEJoin [2], This new physical operator provides a fast algorithm for joins containing only inequality conditions.

REFERENCES:

[1] Z. Khayyat, I. F. Ilyas, A. Jindal, S. Madden, M. Ouzzani, P. Papotti, J.-A. Quian-Ruiz, N. Tang, and S. Yin. BigDansing: A System for Big Data Cleansing. In ACM SIGMOD, pages 1215-1230, 2015.

[2] Z. Khayyat, W. Lucia, M. Singh, M. Ouzzani, P. Papotti, J.-A. Quiane-Ruiz, N. Tang, and P. Kalnis. Lightning Fast and Space Efficient Inequality Joins. PVLDB 8(13): 2074-2085, 2015.

The Web is barely 25 years old but in that time it has changed every aspect of our lives. Because of its sociotechnical rather than purely engineered nature, not only is the Web changing society but also we shape the way the technology evolves. The whole process is inherently co-constituted and as such its evolution is unlike any other system. In order to understand how the Web might evolve in the future - for good or bad - we need to study how it has evolved since its inception and the associated emergent behaviours. We call this new research discipline Web Science [1,2], and it is important for all our futures that we urgently address its major research challenges.

We are fast becoming part of a world of digital inter-connectivity, where devices such as smartphones, watches, fitness trackers, and household goods are part of a growing network, capable of sharing data and information. Increasingly, the Web has become the ubiquitous interface to access this network of devices. From sensors, to mobile applications, to fitness devices, these devices are transmitting their data to - often - centralised pools of data, which then become available via Web services. The sheer scale of this data leads to a rich set of high-volume, real-time streams of human activity, which are often is made publicly consumable (potentially at a cost) via some API. For academia, the combination of these sources are providing social scientists and digital ethnographers a far richer understanding of society, and how we as individuals operate.

These streams represent a global network of human and machine communication, interaction, and transaction, and with the right analytical methods, may contain valuable research and commercial insights. In domains such as health and fitness for example, the aggregation of data from mobile devices is supporting the transition towards the quantified-self, and offers rich insight into the health and well-being of individuals, with the potential of diagnosing or decreasing disease.

Why the need for Web Observatories?

Studying the Web provides us with critical insights about how we as individuals and society operate in the digital world. The actions, communications, interactions, and transactions produced by humans and machines has the potential to offer rich insight into our world, allowing us to better understand how we operate at a micro and macro scale. However, there are a number of barriers that prevent researchers from making the most of those data resources.

Herein lies a challenge, and a great opportunity. We are now in a position where the technologies used within the big data processing pipeline are maturing, as are the methods we use to analyse data to provide valuable insights. Yet, overshadowing these benefits are issues of data access, control, and ownership. Whilst the data being produced continue to grow, their availability beyond the walled-gardens of the data holder - whether commercial or institutional - reduce the full potential of analysis envisaged in the big data era.

(a) Datasets are distributed across different domains (b) Metadata about those datasets are not available or are in different vocabularies/formats, (c) Searching or for or inside datasets is not possible, (d) Applying analytics on one or more datasets requires copying them into a central location, (e) Datasets are often provided in the context of specific disciplines lacking the metadata and enrichment that could make them available in other disciplines, (f) The nature of some of the datasets often requires access control in the interest of privacy. (g) There is a need for engines that lower the barrier of engagement with analytics for individuals, organisations and interdisciplinary research communities by supporting the easy application of analytics across datasets without requiring them to be copied into a central location, (h) There is a need for services enabling the publication and sharing of analytical tools within and across interdisciplinary communities.

Addressing the challenges described above, we have introduced the Web Observatory [3], a globally distributed infrastructure that enables users to share data with each other, whilst retaining control over who can view, access, query, and download their data. At its core, a Web Observatory comprises of a list of architectural principles that describe a scalable solution to enable controlled access to heterogeneous forms of historical and real-time data, visualisations, and analytics. In order to handle these new forms of big, and small data, significant effort has gone into developing technologies capable of storing, querying, and analysing high-volume datasets - or streams - in a timely fashion, returning useful insights of social activity and behaviour.

A Global Network of Web Observatories

The Web Observatory (WO) project, developed under the auspices of the Web Science Trust, aims to develop the standards and services that will interlink a number of existing or emergent Web Observatories to enable the sharing, discoverability and use of public or private datasets and analytics across Web Observatories, on a large, distributed scale (http://online.liebertpub.com/doi/abs/10.1089/big.2014.0035). It involves the publication or sharing of both datasets and analytic or visualisation tools (http://webscience.org/web-observatory/). At the same time, it involves the development of appropriate standards to enable the discovery, use, combination and persistence of those resources; effort in the direction of standards is already underway in the W3C Web Observatory community group (http://www.w3.org/community/webobservatory/).

International research collaboration is one of the primary goals of creating a network of Web Observatories. There has already been significant effort in creating a number of Web Observatory nodes globally [4,5]

In this paper will describe an instance of the Web Observatory, the Southampton Web Observatory (SUWO) and how it is being applied both at Southampton and at other institutions in areas such as integrated health management in particular in support of an aging population

We believe that the true potential of the Web Observatory vision will be realised when the different observatories become part of a global network of Wide Web of Observatories, allowing cross-observatory querying and analysis. Only through working through a set of initial application areas, we will show the immediate value that the Web Observatory platform will provide, from the sharing of datasets and resources, to improving international collaboration and research opportunities as a result of the raised awareness of institutional resources.

References

[1] Berners-Lee, Tim, Hall, Wendy, Hendler, James, Shadbolt, Nigel and Weitzner, Danny Creating a Science of the Web. Science, 313, (5788), 769-771.

[2] Hendler, James, Shadbolt, Nigel, Hall, Wendy, Berners-Lee, Tim and Weitzner, Daniel (Web science: an interdisciplinary approach to understanding the Web. Communications of the ACM, 51, (7), 60-69.

[3] Tiropanis, Thanassis, Hall, Wendy, Shadbolt, Nigel, De Roure, David, Contractor, Noshir and Hendler, Jim, The Web Science Observatory. IEEE Intelligent Systems, 28, (2), 100-104.

[4] Tinati, Ramine, Wang, Xin, Tiropanis, Thanassis and Hall, Wendy, Building a real-time web observatory. IEEE Internet Computing (In Press).

[5] Wang, Xin, Tinati, Ramine, Mayer, Wolfgang, Rowland-Campbell, Anni, Tiropanis, Thanassis, Brown, Ian, Hall, Wendy and O'Hara, Kieron, 2Building a web observatory for south Australian government: supporting an age friendly population. In, 3rd International workshop on Building Web Observatories (BWOW), 10pp.

Mathematicians considered lattices as early as the first half of the nineteenth century, e.g. Johann Carl Friedrich Gauss and Joseph-Louis Lagrange explored point lattices in small dimensions. After the pioneering work of Hermann Minkowski around 1900, lattices were extensively studied in the twentieth century until engineering applications in the areas of digital communications, data compression, and cryptography were recently discovered. Nowadays it is admitted that lattices are good for almost everything, including many new fields such as physical-layer security, post-quantum cryptographic primitives, and coding for wireless mobile channels.

In this talk, after introducing the mathematical background for point lattices in multi-dimensional Euclidean spaces, we shall describe how lattices are used to guarantee reliable and secure communications. The talk will include strong technical material but it is intended to a large audience including engineers and scientists from all areas. The talk shall also present new results on lattices found in Qatar under projects NPRP 6-784-2-329 and NPRP 5-597-2-241 funded by QNRF.

Lattices are mathematical structures with specific algebraic and topological properties. We consider the simplest form of lattices, i.e. lattices in real Euclidean spaces equipped with the standard scalar product. In communication theory, lattices can play different roles in the processing and the transmission of information. They are suitable for vector quantization of analog sources, for channel coding as coded modulations, and also for joint source-channel coding. In the recent literature, lattices are found to be good tools in network coding and secure coding at the physical layer. More information on lattice for communications can be found in [1] and references therein. A lattice is a Z-module of the Euclidean vector space R^N or equivalently it is a discrete additive subgroup of R^N. A new family of lattices, referred to as Generalized Low-Density Lattices (GLD) was built in Qatar. GLD lattices are obtained by Construction A from non-binary GLD codes. Their impressive performance and analysis can be found in [2].

In his seminal paper [3], Miklos Ajtai showed how lattices could be used as cryptographic primitives. Since Ajtai's result, lattice-based cryptography became very popular in the cryptography community. One important area is Learning With Errors (LWE), see [4]. LWE is a clear example of elementary mathematical problem with an algorithmically complex solution. Thanks to its intrinsic connection with lattice-based problems that are known to be hard also for quantum algorithms, LWE has raised much interest in the last decade, it is part of the so-called post-quantum cryptography. Another area in lattice-based cryptography is GGH. This is the first McEliece-like scheme for lattices, referred to as GGH cryptosystem, and was proposed by Goldreich, Goldwasser, and Halevi in 1997. GGH is a lattice equivalent of the McEliece cryptosystem based on error-correcting codes. Many other cryptosystems based on lattices were investigated and proposed by several researchers. In this talk, after presenting how lattices are used for reliable communications, we also present how lattices are used to secure data communications. The attached slides constitute a draft mainly focusing on the communication aspects of lattices. These slides will be completed by a second part on security based on lattices.

[1] R. Zamir, Lattice Coding for Signals and Networks, Cambridge, 2014.

[2] J.J. Boutros, N. di Pietro, and Y.C. Huang, “Spectral thinning in GLD lattices”, Information Theory and Applications Workshop (ITA), La Jolla, pp. 1-9, Feb. 2015. Visit http://www.ita.ucsd.edu/workshop/15/files/paper/paper_31.pdf

[3] M. Ajtai, “Generating Hard Instances of Lattice Problems,” Proc. of the Twenty-Eighth Annual ACM Symposium on Theory of Computing, pp. 99-108. doi:10.1145/237814.237838, 1996.

[4] V. Lyubashevsky, C. Peikert, and O. Regev, “A Toolkit for Ring-LWE Cryptography,” in EUROCRYPT, pp. 35-54, 2013.

This paper presents an abstract of the final phase of an on-going research project aiming at investigating the antecedents and consequences of research and innovating within Lebanese small and medium-sized enterprises (SMEs). The role of training personnel and introducing supporting activities in moderating the relationship between R & D and introducing knowledge management systems within Lebanese small- medium –sized innovation in Lebanon still suffer from funding shortages, short of IT personnel training and lack of the ability to adequately use existing knowledge. Ashrafi, and Murtaza suggest that “Large organizations have enough resources to adopt ICT while on the other hand SMEs have limited financial and human resources to adopt ICT” (Ashrafi and Murtaza, 2008 P. 126). Even though, Lebanese government is trying to create a digital economy, Lebanon ranked 94th out of 144 countries on the Network Readiness Index in 2012 and “In the Arab world, Lebanon ranked in 10th position, right behind Morocco (89th worldwide), but right ahead of Algeria (131st Worldwide)” (BankMed, 2014; P.19). What it is imperative, however, to note here is that “SMEs have been recognized as an important source of innovative product and process” (HanGyeol et al 2015; P.319). It is widely believed that “Research and development (R&D) intensity is crucial for increasing the innovative capacity of small to medium-sized enterprises (SMEs)” (Nunes et al 2010; P.292). Most interestingly, “the Labour Market Survey (2001) showed a clear relationship between business failure and a lack of planning or training by SMEs” (Javawarna et al 2007; P.321). Based on existing review of literature, it is found that business expansion oblige SMEs to adopt new and original information technology solutions. Simultaneously, it is suggested that “Lack of training and skills of IT in organizations will result in a limited use of IT and lack of success in reaping benefits from computer hardware and software in organizations” (Ghobakhloo et al 2012, P.44). All that is said is true, “Information technologies (IT) have become one of the most important infrastructural elements for SMEs” (Uwizeyemungu and Raymond, 2011 P.141). As a result, it is generally believed that information technology has an imperative role to play in gaining innovation and competitiveness for SMEs. What is more, it has always been recognized that investing in technology is necessary but insufficient by itself. An imperative need exists for businesses of all sizes to protect their customers by protecting themselves from cyber attack. This is to be accomplished via changing the attitude towards cyber security and the development of cyber culture. Valli and his associates note that “There is little literature available about the ability of SMEs to deploy, use and monitor cyber security countermeasures” (Valli et al 2014;P. 71). Borum and his colleagues believe that “Industries and commercial sectors must collaborate with the government to share and disseminate information, strengthen cyber intelligence capabilities and prevent future cyber incidents” (Borm et al 2015 P.329). Uwizeyemungu and Raymond suggest that “IT adoption and assimilation in these firms should be the product of an alignment between the strategic orientation and competencies that characterize their organizational context on one hand, and specific elements in their technological and environmental contexts on the other hand” (Uwizeyemungu and Raymond 2011, P.153). A study by Ghobakhloo and his colleague “suggested that through the passing of cyber laws by governments to regulate and secure online transaction activities, and also by providing appropriate anti-virus and/or firewall/security protocols for SMEs by vendors and service providers to reduce or prevent the attacks of hackers, viruses and spyware, the perceived risk of IT adoption by these businesses, should be alleviated” (Ghobakhloo et al 2012, P.57). To lead the way to successful innovation within SMEs, this study will be a significant effort in promoting the key sustainability issues affecting innovation within SMEs. The best way to start is by understanding innovation within the Lebanese SMEs. A study such as the one conducted here is recommended by experts in this area. Armbruster and his associates noted that “There is still plenty of research to do before organizational innovation surveys achieve the degree of homogeneity and standardization that advanced R&D and technical innovation surveys possess” (Armbruster et al 2008, P.656). The purpose of this investigation is to determine the relative importance of introducing supporting activities, training personnel and R & D on the variation in introducing knowledge management systems within Lebanese SMEs. To this end, The aim of this project is to investigate the adoption of existing technologies to new applications in a concrete SME business case in addition to what motivate innovation within Lebanese SMEs and the challenges and barriers facing SMEs in adopting innovation. The population of the study consists of all SMEs in Lebanon. Most of SMEs are family businesses and as to be expected “Family involvement in a firm has an impact on many aspects of organizational behaviour” (Cromile and O'Sullivan, 1999, p. 77). Morris and his colleagues argue that “family firms violate a tent of contemporary models of organizations, namely, the separation of ownership from management” (Morris et al, 1996, p. 68) This leads to a lot of complications including the succession problems, role conflict and role ambiguity that may represent major barriers to adopting IT and innovation within SMEs. The sample for this study is relatively large sample and the instrument for collecting the primary data was a well constructed questionnaire. Cronbach's alpha and factor analysis were used to establish the reliability and construct validity of the instrument. Findings of the study show that introducing new or significantly improved supporting activities, training personnel, having an employee who is fully in charge of the website and having an R&D department are the significant factors affecting introducing new or significantly improved knowledge management systems to better use or exchange information, knowledge and skills within Lebanese SMEs. Findings of this study are in line with previous findings. Schienstock and associates believe that firms have to develop their competence to learn and innovate by introducing new knowledge management practices and organizational restructuring. In fact, they criticized the traditional approach in the classical studies of “the so-called linear model, traditional innovation policy focuses primarily on the creation of new scientific and technical knowledge, supposing some kind of automatic transformation of this new knowledge into new products” (Schienstock et al 2009, PP.49–50). What is more, Molero and García believe that “the theory about factors affecting firms' innovation has still a long way to go because the analytical object is complex and difficult to set limits for” (Molero and García 2008, P.20). This project has implications for policy making, decision making and recommendations for further research.

References:

Armbruster, H. Bikfalvi, A. Kinkel, S. and Lay, G (2008) “Organizational innovation: The challenge of measuring non-technical innovation in large-scale surveys”, Technovation 28, 644–657. The full text is available at: www.sciencedirect.com

Ashrafi, R. and Murtaza, M. (2008), “Use and Impact of ICT on SMEs in Oman.” The Electronic Journal Information Systems Evaluation Volume 11 Issue 3, pp. 125–138. The full text is available at: www.ejise.com

BankMed (April 2014) “ANALYSIS OF LEBANON'S ICT SECTOR”, P.19. The full text is available at: http://www.bankmed.com.lb/LinkClick aspx?fileticket = xGkIHVHVrM4%3D&portalid = 0

Borum, R.; Felker, J.; Kern, S.; Dennesen, K; Feyes, T. (2015), “Strategic cyber intelligence”, Information & Computer Security, Vol. 23 Iss: 3, pp.317–332. The full text is available at: http://www.emeraldinsight.com.ezproxy.aub.edu.lb/doi/10.1108/ICS-09-2014–0064

Cromile, S. and O'Sullivan, S. (1999), “Women as managers in family firms”, Women in Management Review, Vol. 14 No. 3, pp.76–88. The full text is available at: http://www.emeraldinsight.com/doi/abs/10.1108/09649429910269884

Ghobakhloo, M.; Hong, T.S.; Sabouri, M.S.; Zulkifli, N. (2012), “Strategies for successful information technology adoption in small and medium-sized enterprises”. Information 3, 36–67. The full text is available at: file:///C:/Users/PC/Desktop/information-03-00036%20(2).pdf

Hangyol, S. Yanghon, C. Dongphil, C and Chungwon, W. (2015), “Value capture mechanism: R&D productivity comparison of SMEs”, Management Decision, Vol. 53 Iss: 2, pp.318–337. The full text is available at: http://www.emeraldinsight.com.ezproxy.aub.edu.lb/doi/abs/10.1108/MD-02-2014–0089

Javawarna, D., Macpherson, A. and Wilson, A. (2007), “Training commitment and performance in manufacturing SMEs: Incidence, intensity and approaches”, Journal of Small Business and Enterprise Development, Vol. 14 Iss: 2, pp.321–338. The full text is available at: http://www.emeraldinsight.com.ezproxy.aub.edu.lb/doi/abs/10.1108/14626000710746736

Molero, J. and García, A. (2008), “Factors affecting innovation revisited”, WP05/08, PP:1–30. The full text is available at: https://www.ucm.es/data/cont/docs/430-2013-10-27-2008%20WP05-08.pdf

Morris, M. H. Williams, R. W. Nel, D. (1996), “Factors influencing family business succession”, International Entrepreneurial Behaviour & Research. Vol. 2 No. 3, pp.60–81. The full text is available at: http://www.emeraldinsight.com/doi/abs/10.1108/13552559610153261

Nunes, P.M. Serrasqueiro, Z. Mendes, L. Sequeira, T.N (2010), “Relationship between growth and R&D intensity in low-tech and high-tech Portuguese service SMEs”, Journal of Service Management, Vol. 21 Iss: 3, pp.291–320. The full text is available at: http://www.emeraldinsight.com.ezproxy.aub.edu.lb/doi/pdfplus/10.1108/09564231011050779

Schienstock, G. Rantanen, E. and TyniIAREG, P. (April 2009) “Organizational innovations and new management practices: Their diffusion and influence on firms' performance. Results from a Finnish firm survey”. IAREG Working Paper 1.2.d. PP. 1–64. The full text is available at: http://www.iareg.org/fileadmin/iareg/media/papers/WP_IAREG_1.2d.pdf

Uwizeyemungu, S. and Raymond, L. (2011), “Information Technology Adoption and Assimilation: Towards a Research Framework for Service Sector SMEs,” Journal of Service Science and Management, Vol. 4 No. 2, pp. 141-157, doi: 10.4236/jssm.2011.42018. The full text is available at: http://www.scirp.org/Journal/PaperInformation.aspx?PaperID5229 World Economic Forum (2015), Global Information Technology Report 2015. The full text is available at: http://www3.weforum.org/docs/WEF_Global_IT_Report_2015.pdf

Valli, C. Martinus, I. and Johnstone, M. (Aug 2, 2014), “Small to Medium Enterprise Cyber Security Awareness: an initial survey of Western Australian Business”, The 2014 International Conference on Security and Management, At Las Vegas, Nevada, PP:71-75. The full text is available at: https://www.researchgate.net/publication/264417744_Small_to_Medium_Enterprise_Cyber_Security_Awareness_an_initial_survey_of_Western_Australian_Business

The United Arab Emirates (UAE) is the eighth largest oil producing country in the world. It has about 800 km of coastline. Beyond the coastline, its territorial water and exclusive economic zone have very rich and extensive marine life and natural resources. Most of the oil and natural gas in the UAE is produced from the offshore oil fields. Maritime oil exploration and transportation has increased more steeply due to the expansion of the world crude oil and natural gas production, and the trend of using larger-shaped and higher-speed container vessels. The probability of oil rigs pollution, burning and explosion continues to rise. All these factors stimulate a greater danger for vessels, oil operation safety and maritime environment. Therefore, maritime security and environmental protection are of great interest, both from the academia and petroleum industry point of view. The continuous surveillance of the offshore oil fields is essential to secure the production flow, avoid trespassing and prevent vandalism from intruders and pirates. With the emergence of the new technologies such as maritime wireless mesh network (MWMN) and camera sensor network (CSN), maritime surveillance systems have been gradually improving due to the accuracy, reliability and efficiency of the maritime data acquisition systems. However, in order to realize the oil operation security, it is necessary to control and implement a dynamic system to monitor the maritime environment. The monitoring objects include vessels, fishery boats, pollution, navigational and sailing conditions. By the same token, the legacy monitoring systems such as very-high frequency (VHF) communication, marine navigational radar, vessel traffic service (VTS) and automatic identification system (AIS) are still insufficient to satisfy the increasing demand of the maritime surveillance. The objective of the paper is to provide full-view coverage CSN for a triangular grid based deployment and to reduce the total transmission power of the image to cope with the limited available power in the CSN. The rough and random movements of the sea surface can lead to a time-varying uncovered area by displacing the CSN from its initial location. Thus, it is important to investigate and analyze the effects of the sea waves on the CSN to provide full-view coverage in such complex environments. The main challenges in the deployment of the CSN are the dynamic maritime environment and the time-varying full-view coverage provided by the sea waves. Therefore, quasi-mobile platforms such as buoys are envisaged to hold the camera sensor nodes. The buoys will be nailed at the sea floor to limit the movement of the buoys due to the sea waves. In addition, cooperative transmission method has been proposed to reduce the total transmission power of the image in the CSN. A CSN is formed by autonomous, self-organized ad-hoc camera sensor nodes that are equipped with wireless communication devices, processing unit and power supply. The design, implementation and deployment of a CSN for maritime surveillance stimulate new challenges different to that which exist on the land, as the maritime environment hinders the development of such a network.

The main differences are summarized as follows:

• Dynamic aspect of the maritime environment which requires rigorous levels of device protection.

• Deployment characteristic of a CSN in maritime environment which is highly affected by wind direction and speed, sea wave, and tide.

• Requirement of flotation and anchoring platforms for a CSN and the possible vandalism from intruders and pirates.

• Coverage problem of a CSN due to the random and rough sea movement.

• High energy consumption if battery-power based cameras are used continuously.

• Communication signals are highly attenuated by the constant sea movement.

In this context, CSNs with ubiquitous and substantive camera sensor nodes can be utilized to monitor offshore oil fields to secure the production flow, avoid trespassing and vandalism from intruders and pirates. However, camera sensor nodes can generate various views of the same target if they are captured at different viewpoints if the image is taken near or at a frontal viewpoint, then, the target will be more likely to be recognized by the recognition system. It is fundamental to understand how the coverage of a given camera depends on different network parameters to better design numerous application scenarios. The coverage of a particular CSN represents the quality of surveillance provided by the CSN. As the angle between the target's facing direction and the camera's viewing direction increases, the detection rate drops severely. Consequently, the camera's viewing direction has a considerable effect on the quality of surveillance in a CSN. Recently, a novel concept, which is called full-view coverage, has been introduced to characterize the intrinsic property of camera sensor nodes and assess the coverage in CSNs. A target is full-view covered if its facing direction is always within the scope of a camera, regardless of the target's actual facing direction. Simply, the underlying contribution of full-view coverage tackles the pledge of capturing the target's face image. Consequently, designing a CSN with full-view coverage is of major importance, as the network does not only provide the detection of a target, but also the recognition of it. In many network configurations, camera sensor nodes are not mobile and they remain stationary after the initial deployment. In a stationary CSN, when the deployment characteristic and sensing model for the CSN are defined, the coverage can be deduced and remain unchanged over time. In order to address the hostile maritime environment, there has been a strong desire to deploy sensors mounted on quasi-mobile platforms such as buoys. Such quasi-mobile CSNs are extremely beneficial for offshore oil field surveillance where buoys move with the sea waves. Hence, the coverage of a quasi-mobile CSN depends not only on the initial network deployment, but also on the mobility pattern of the CSN. Nevertheless, the full-view coverage under a quasi-mobile CSN in maritime network has not been investigated. This problem is pivotal for the network design parameters and application scenarios of CSNs where conventional deployment characteristics such as air-drop fail or is not appropriate in a maritime environment. Since a priori knowledge of the terrain is available, a grid-based deployment can be utilized for the given terrain. The endeavor to design a practical mobility pattern for CSNs gives rise to model the cable attached to the buoy as a spring. In this practical mobility pattern, buoys start from an initial coordinate assignment, then oscillate based on spring force, sea wave, wind direction and speed and ultimately converge to a consistent solution. Specifically, this mobility pattern is based on two stages. The first stage is the effects of sea wave, wind direction and speed that move a buoy. The second stage is the spring reaction, based on the previous effects. Then, the design concept is followed and extended to develop a mobility pattern for CSNs in maritime environment. A CSN is considered that constitutes of a small number of buoys whose locations are initially known and consecutively their locations will be derived based on spring relaxation technique. With this technique, coverage issues arisen in a CSN and design a cooperative transmission method to reduce the total transmission power in the CSN are studied. One primary problem is how to design a realistic sea wave model for a given deployed CSN to achieve full-view coverage. Compared with the traditional sea wave model which assumes sine wave model for simplicity analysis, there are two elements that increase the complexity of the problem in a realistic sea wave model. First, the force that acts on the surface of the sea, which is supposed to be the main driving force for the creation of waves in deep water. Second, the force from the sea surface and ground interaction, which is the main contributing force near shoreline, however, this type of force becomes dominant in deep water when seaquakes occur. The realistic sea wave model should encounter the extruding of a two dimensional sine wave model into a three dimensional sea wave model. However, there should be some order of variation along the wave propagation direction for the finite wide waves. In conventional wireless sensor networks (WSNs), scalar phenomena can be traced using thermal or acoustic sensor nodes. In camera sensor networks (CSNs), images and videos can significantly enrich the retrieved information from the monitored environment, and hence provide more practicality and efficiency to WSNs. Recently; there has been enormous development of applications in surveillance, environment monitoring and biomedicine for CSNs that has brought a new spectrum to the coverage problem. It is indispensable to understand how the coverage of a camera depends on various network parameters to better design numerous application scenarios. In many network configurations, cameras are not mobile and they remain stationary after the initial deployment. However, different from a stationary CSN, maritime environment poses challenges on the deployment characteristic and mobility pattern for CSNs. In stationary CSNs, when the deployment characteristic and sensing model are defined, the coverage can be deduced and remain unchanged over time. In the maritime environment, camera sensors are mounted on quasi-mobile platforms such as buoys. This paper aims to provide full-view coverage CSN for maritime surveillance using cameras mounted on buoys. It is important to provide full-view coverage because in full-view coverage, targets facing direction is taken into account to judge whether a target is guaranteed to be captured. Image shot at the frontal viewpoint of a given target considerably increases the possibility to detect and recognize the target. The full-view coverage has been achieved using equilateral triangle grid-based deployment for the CSN. To accurately emulate the maritime environment, a mobility pattern has been developed for the buoy which is attached with a cable that is nailed at the sea floor. The buoy movement follows the sea wave that is created by the wind and it is limited by the cable. The average percentage of full-view coverage has been evaluated based on different parameters such as equilateral triangle grid length, sensing radius of camera, wind speed and wave height. Furthermore, a method to improve the target detection and recognition has been proposed in the presence of poor link quality using cooperative transmission with low power consumption. In some parameter scenario, the cooperative transmission method has achieved around 70% improvement in the average percentage of full-view coverage of a given target and total reduction of around 13% for the total transmission power PTotal(Q).

I. Abstract: Despite recent advances in sensor and mobile technology, there is still a lack of an accurate, scalable, and non-intrusive way to knowing how much sunlight we are exposed to. For the first time, we devise a mobile phone software application (SUN BATH), that utilizes a variety of on-board sensors and data sets to accurately predict the sunlight exposure each person is exposed to. The algorithm is able to take into account the sunlight exposure based on the person location, the local weather, sun location, and shadow from buildings. The algorithm achieves this by using the mobile user's location and other sensors to determine whether it is indoors or outdoors, and uses building data to calculate shadow effects and weather data to calculate diffused light contributions. This will ultimately allow the user to be more informed about sunlight exposure and compare it with daily recommended levels to encourage positive behaviour change. In order to show the value added by the application, SUN BATH is distributed to a sample of students population for benchmarking and user experience trials. The latest stable version of the application, suggests a scalable and affordable way compared to survey or physical sensing methods. II. Introduction: In this particular proposal, we examine how to live healthily in cities using a data-driven mobile-sensing approach. Cities are partly defined by a high building concentration and a human lifestyle that is predominantly spent indoors or in the shadow of buildings. Some cities in particular, also suffer from heavy pollution effects that significantly reduces the level of direct solar radiation. As a result, one area of concern is the urban dwellers lack of exposure to the ultra-violet (UV) band of sunlight and the wide range of associated health problems. The large scale and chronic nature of the health problems can lead to a time bomb in the National Health Service and cause irreversible future damage to the economy. This article proposes using the ray tracing SORAM model by Erdelyi et al. as an innovative and flexible technique for modelling and estimating the amount of solar irradiation can be collected at a time and certain location. SORAM module is already benchmarked against real measurement data, hence, our work here will benefit from this by taking the calculated ray-tracing information as a primary filter. The aim is to devise an affordable and accurate way of continuously estimating each person's UV exposure. Primarily, this is achieved by developing an Android smartphone application that uses the SORAM advanced modelling techniques to estimate the level of UV exposure each person is subjected to at any given time and location. The research novelty is that the proposed solution does not require additional purpose-built hardware such as a photovoltaic sensor, but instead utilizes a combination of accurate wireless localization, and weather-/terrain-informed sunlight propagation mapping. The challenges addressed include how to accurately locate a human and how to model the propagation of sunlight in complex urban environments. The latest stable version of the application, suggests a novel and affordable way compared to traditional or physical methods when calculating the amount of sunshine we are exposed to. III. System Overview: We implemented and evaluated SUN BATH application with the Android platform using different mobile phone models such as Samsung S5, Asus Zen5 and Archos tablet. The application is developed using Android Studio as IDE for Android application development. The application allows the user to create a profile using a user name and some information such as date of birth, height, weight, skin colour, country of origin, and level of income. To be used later for future detailed reporting with relation to the amount of sun bathing for different groups and ethnicities. SUN BATH only relies on lightweight “sensors to server” modelling which allows continuous low-energy and low-cost tracking of the user location and state transitions. In particular, we will present the process to show that we were able to use SORAM within the smart-phone environment to accurately infer the amount of sunshine in a user is exposed to based on the accuracy of the GPS and other location modules for Android mobile phones. To meet stringent design requirements, SUN BATH utilizes a series of lightweight ‘sensors – server’ for a fault-tolerant location detection. SUN BATH primarily makes use of three types of location-aware detectors: WiFi, cellular-network, and GPS. The aforementioned three wireless location detectors are used in conjunction to improve resolution and resilience. WiFi hub SSID identifiers are used to locate the hub in known open and commercial databases up to an accuracy of a few metres. In the absence of WiFi, a combination of cell tower location area and assisted GPS is used to get an accuracy of 10–15 m in urban areas with shadow effects. WiFi detector adopts the distributed IP address to capture the source location to determine the region the user is in. Cellular-network detector detects the source and attenuation of signals caused by objects on its path (e.g., trees, buildings). It normally help to indicate the movement of the user as the mobile signal gets handed from one network to another. The Application utilizes the GPS sensor to exactly pinpoint the coordinates of the user location i.e. Latitude and Longitude. The system time clock is also used to assist the detection of the local time. The App cache-in those parameters and sends it to a remote server whenever there is an Internet connection. The server hosts the SORAM calculation algorithm which generates a live estimate of the amount of sun exposure the user is experiencing. The results are then passed back to the applications through the Open Database Connectivity (ODBC) middleware service to permanently store the results in a secure database management system (DBMS). IV. Soram Ray-tracing Methodology: A person positioned in an out-door environment is surrounded by solar radiation, which consists of direct and diffuse rays. Direct and diffuse radiation data on a horizontal surface are usually collected at various locations and weather stations across the universe. The raw datasets collected can be used to estimate the amount of global radiation at any point of earth of a given slope and azimuth. Due to cost and scarcity of live data, the SORAM algorithm embed and made use of the Reindel Model, to estimate the direct and diffuse irradiation from hourly horizontal global radiation data. In addition and to go light on computation and avoid calculations for the nighttime hours, the SORAM determines the sunrise time for each day of the year and the amount of solar radiation data from that point onwards which is then calculated until sunset. The algorithm also estimates and with high accuracy direct and diffuse radiation on a surface of given slope and azimuth from their counterparts on a horizontal access considering surrounding shading conditions. We tested SUN BATH in simulated and real locations for five continuous days from sunrise to sunset in around the School of Engineering building complex at the University of Warwick campus. Simulated tests were carried manually, using two fixed location parameters i.e Longitude and Latitude. A quick memory and CPU monitor view revealed that the SUN BATH energy consumptions and resource-constraint on the used smartphone devices were moderate. A full memory and CPU monitor view can be easily produced, but it is beyond the scope of this article. V. Conclusions and Future Work: This research presented the architecture of SUN BATH mobile sensing application that gathers a variety of lightweight sensors information and utilised ray-tracing algorithm to derive the level of human sun exposure in urban areas. The application has demonstrated that it can be an affordable and pervasive way of accurately measuring the level of sunlight exposure each person is exposed to. Further work is required to scale the project to the global level, which requires big data sets on urban building maps and meterological data from all the cities.

Patient Engagement's Research results a mobile Application Patient Center to Manage Appointments, Personal Health Records and to avail Emergency Online Healthcare across all hospitals in Patient Center Network. It's Single Mobile App to manage and engage Patients actively in Healthcare across Qatar.

Problem Statement: Healthcare trends moving towards Patient Centric Healthcare Model from Hospital Centric Model. Qatar is having more than 350 Health facilities. Most of Health facilities not connected each other. Global Healthcare trends are implementing Interoperable Systems in Hospitals (Integrated Health Network), so that they can exchange their health data and Patient's Care will be continued and data will be shared with partnered networks and also government entities Biggest Gap in Healthcare process will be Patient Engagement. Providers and patients working together to improve health. A patient's greater engagement in healthcare contributes to improved health outcomes, and information technologies can support engagement. Patients want to be engaged in their healthcare decision-making process, and those who are engaged as decision-makers in their care tend to be healthier and have better outcomes Outpatients in Hospitals. Outpatient's work flow is very much defined in Healthcare Process, By Current technology, we need to reduce out patient's visit to hospital by having remote healthcare facility, so that Providers can provide better health outcomes on Inpatients Study: Book Name: Engage! Transforming Healthcare Through Digital Patient Engagement Edited by Jan Oldenburg, Dave Chase, Kate T. Christensen, MD, and Brad Tritle, CIPP This book explores the benefits of digital patient engagement, from the perspectives of physicians, providers, and others in the healthcare system, and discusses what is working well in this new, digitally-empowered collaborative environment. Chapters present the changing landscape of patient engagement, starting with the impact of new payment models and Meaningful Use requirements, and the effects of patient engagement on patient safety, quality and outcomes, effective communications, and self-service transactions. The book explores social media and mobile as tools, presents guidance on privacy and security challenges, and provides helpful advice on how providers can get started. Vignettes and 23 case studies showcase the impact of patient engagement from a wide variety of settings, from large providers to small practices, and traditional medical clinics to eTherapy practices. Book Name: Applying Social Media Technologies in Healthcare Environments Edited by Jan Oldenburg, Dave Chase, Kate T. Christensen, MD, and Brad Tritle, CIPP Applying Social Media Technologies in Healthcare Environments provides an indispensable overview of successful use of the latest innovations in the healthcare provider-patient relationship. As professionals realize that success in the business of healthcare requires incorporation of the tools of social media into all aspects of their worlds and recognize the value offered by the numerous media channels, this compendium of case studies from various voices in the field – caregivers, administrators, marketers, patients, lawyers, clinicians, and healthcare information specialists – will serve as a valuable resource for novices as well as experienced communicators. Written by experienced players in the healthcare social media space and edited with the eye of an administrator, chapters provide insight into the motivation, planning, execution, and evaluation of a range of innovative social media activities. Complete with checklists, tips, and screenshots that demonstrate proven application of various social channels in a range Based on the research on Patient Engagement, I have designed mobile based Patient Engagement solution called Patient Center. Using Patient Center mobile app, we can search nearest Specialties, Doctors with review ratings and also can book appointments in Health Facilities which are part of Patient Center Network. Patients can securely access their health records mobile and also can exchange health records to any provider for better treatment. Patient Center App works as personal health advisor by reminding patients medications, exercises and discharge notes etc. Patient Center is trying solve problems in following areas Reducing Phone calls for booking appointments and reducing time to fill forms prior to Encounter Patient Center Network contains several Providers, Payers, Pharmacies and Laboratories, all are completely interoperable and exchange Health Data with HIPPA Compliance Patients no need to carry bundles of his/her health records in papers when they meet doctor, instead all personal health records across all hospitals can be downloaded to Patient's Mobile and Patient can securely exchange his/her health data to any provider Single Mobile App to manage and access Appointments, lab orders medications, diagnoses, care plans, immunization history, and more On Emergency Cases, Patient Can click Emergency Service Menu in App and automatically nearest ambulance services will be alerted with Patient's GPS Location Complex Discharge Notes will be illustrated in Mobile App with images. Medication Consumption Alarms as per doctor's Advice in mobile Communications with Doctor from home using Patient Center, Healthcare Wearables which can provide Heart Rate, Blood Pressure, ECG, Diabetes Readings to Doctor in Real Time, so that, Doctor can diagnose patient online. Patient Center will try to reduce 60% outpatient visits and get Quality Healthcare right from Home In addition to the resources above, Patient Center will focus on following areas Social and Behavioral which covers social media, texting and gaming, wearables and mobile, and the social determinants of health Home Health which covers remote monitoring and telehealth, patient education, and smart homes Financial Health which includes managing health insurance and expenses, transparency and consumerism, patient onboarding and financial options Technology We are going to build entire application using IHE Profiles to achieve Interoperability in our networked hospitals (Integrated health Management). Integrated Health Enterprise (IHE) is an initiative by healthcare professionals and industry to improve the way computer systems in healthcare share information. IHE promotes the coordinated use of established standards such as DICOM and HL7 to address specific clinical needs in support of optimal patient care. Systems developed in accordance with IHE communicate with one another better, are easier to implement, and enable care providers to use information more effectively. The Exchange of Personal Health Record Content (XPHR) profile provides a standards-based specification for managing the interchange of documents between a Personal Health Record used by a patient and systems used by other healthcare providers to enable better interoperability between these systems. The Exchange of Personal Health Record Content (XPHR) integration profile describes the content and format of summary information extracted from a PHR system used by a patient for import into healthcare provider information systems, and vice versa. The purpose of this profile is to support interoperability between PHR systems used by patients and the information systems used by healthcare providers. Patient Center leverage other IHE Integration and Content Profiles for interoperability in addition to the XPHR Content Profile. For example, a PHR Systems may implement XDS-MS to import medical summaries produced by EHR systems, XDS-I to import imaging information, XDS-Lab to import laboratory reports, etc Patient Center Mobile Application is connected to Intersystems Healthshare Platform which is Health informatics platform certified by IHE. Patient Center app will be developed in ios and Android native application development platforms and connected to Healthshare server using web services and REST. Healthshare will be connected to network hospitals using IHE transactions. We use IHE XPHR for personal health records exchange.

Conclusion

These scenarios are not theoretical; early adopting consumers across these various patient personae are currently using Patient Engagement Solutions. Consumers appreciate the idea of remote health monitoring, and 89% of physicians would prescribe a mobile health app to patients. Patients continue to trust their personal physicians above most other professionals, following nurses and pharmacists. Patient Center is designed for better Health outcomes, Clinical Decision Support, Engaging Patients actively in Healthcare and reducing outpatient visits and repetitive encounters.

I. Introduction

Studies reveal that often group members collaborate when searching for information even if they were not explicitly asked to collaborate [1]. The activity that involves a group of people engaging in a common information seeking task is called Collaborative Information Seeking (CIS). Over the past few years, CIS research has focused on providing solutions and frameworks to support the process [2]. However, work in this field to date has always assumed that information seekers engaged in CIS activity are using the same access modality, the visual modality. The attention on this modality has failed to address the needs of users who employ different access modalities such as haptic and/or audio. Visual Impaired (VI) employees in a workplace may often have to collaborate with their sighted team members when searching the web. Given that the VI individual's search behaviour is challenged by poor web design and the shortcomings of current assistive technology [3][4]; collaboratively engaging in web search activity with peers can be considered as a major barrier to workplace inclusion. This study explores the under investigated area of cross-modal collaborative information seeking (CCIS), that is the challenges and opportunities that exist in supporting VI users to take an effective part in collaborative web search tasks with sighted peers in the workplace.

II. Study Design

To develop our understanding of the issues, we investigated the CCIS activity of pairs of VI and sighted participants. The study consisted of an observational experiment that involved 14 VI and sighted users completing two web search tasks which was followed up by scenario-based interviews which conducted with seven of the 14 pairs from the experiment. We conducted the experiment to examine the patterns of CCIS behaviour and the challenges that occur. In the scenario-based interviews we examined the techniques used, tools employed and the ways information is organized both for individual and collaborative use. In the observational study, all VI participants used a speech-based screen reader. Each pair was given two search tasks; one task was performed in a co-located setting and the other task was performed in a distributed setting. For the co-located task, the participants were asked to work collaboratively to organize a trip to the United States while for the distributed task they were asked to plan a trip in Australia. In the distributed condition, participants were seated in different locations and were told that they were free to use any method of communication they preferred; 5 pairs used emails and 9 pairs used Skype. Seven pairs from the observational studies took part in the scenario-based interviews. The interviews involved the interviewer describing a CIS activity to the participants followed by four scenarios that contain questions relating to the management of the retrieved information.

III. Observational Study Findings

A. Division of Labour

In the co-located condition, discussion about the division of labour occurred at two levels. Firstly in the initial discussion and secondly as a result of one participant interrupting his/her partner in order to complete a certain action. Three reasons were identified for these interruptions: (1) When viewing large amounts of information with assistance from their sighted participant. (2) When browsing websites with inaccessible components, VI users asked sighted participants to perform the task or assist them in performing it. (3) The third reason is related to the context of the task. In contrast, in the distributed condition, the discussion about the division of labour occurred only at the beginning of the task. The pair divided the work and started work independently. The participants only updated each other about their progress through the communication tool. Unlike, the co-located sessions, collaboration in the later stages was not observed. Additionally, VI participants' requests for assistance were fewer in this condition as they seemed to be more reluctant to ask for support in the distributed condition. When a VI participant encountered an accessibility issue, they would try on average 3 websites before asking their sighted partner to assist them. The majority (13 pairs in the co-located and 12 pairs in the distributed setting) divided the labour so that sighted participants performed booking related activities and VI participants performed event organization activities. VI participants emphasised that they chose this approach to avoid any issues related to accessibility. Vigo and Harper [5] categorized this type of behaviour as “emotional coping”, in which users past experience relating to an inaccessible action in a similar webpage or similar task affects their judgment on website use or tasks conducted. It is clear from the results that VI users put some thought into either dividing the labour in a specific way or to find some other way to get around the issues encountered.

B. Awareness

In the co-located condition, the main method to implement awareness was verbal communication. In the distributed condition, the only method to implement awareness was through email and instant messaging. To facilitate awareness of partner's activities while performing the task, participants informed their partners about actions. The participants tended to provide their partner constantly with information about their current activities to enrich group awareness in the absence of a tool that supported awareness in both conditions. In fact, pairs who completed more of the task in both conditions communicated more information about their activities. It was observed that when more information to avoid duplication of effort was communicated, the higher was the performance in the distributed condition. This indicates that making this type of information available between distributed collaborators might enhance their ability to complete tasks efficiently. This was not the case in the sessions in the co-located condition, as the sessions with lowest and highest performance reported the same amount of information communicated relating to duplication of effort. However, it was observed that pairs who performed well in the co-located sessions have communicated more information related to the actions they are performing that is not essential for the other pair to know. This indicates that facilitating the appropriate type and amount of awareness information in each condition is crucial to team performance and can increase team productivity [6].

C. Search Results exploration and management

Collaboration occurred mainly in two stages of the information seeking process: the results exploration and results management. In the results exploration stage, collaboration was triggered when VI participants viewed large amounts of information with their sighted partner's assistance or by both partners deciding to explore search results together. The average number of search results viewed collaboratively was higher than the average number of search results viewed by VI participants alone. Screen readers' presentation of large volumes of data imposes a number of challenges such as short term memory overload and a lack of contextual information [3]. This stage is highlighted as one of the most challenging stages faced by VI users during the IS process [4]. The amount of retrieved information kept by sighted users is nearly double the amount of information kept by VI users. The reasons for this were twofold: (1) Sighted users viewed more results than their VI partners. (2) The cognitive overhead that VI users experience when switching between the web browser and an external application used to take notes. This increased cognitive load is likely to slow down the process. The effect of this is more apparent in the distributed condition where VI users are required to switch between three applications: the email client or instant chat application, the web browser and the note taking application.

IV. Scenario-based interviews findings

The scenario-based interview is a tool that allows exploration of the context of the task. This type of scenario narrative approach provides a natural way for people to describe their actions in a given task context. The scenario-based interviews revealed that collaborative searching is quite a common practice as all the participants were able to relate the given scenario with similar activities they had undertaken in the past. It found that often ad hoc combinations of everyday technologies are used to support this activity rather than the use of dedicated solutions. There were clear instances of the use of social networks such as Twitter and Facebook to support the sharing of retrieved results by both VI and sighted interviewees. Individual and cross-modal challenges were also extensively mentioned by VI interviewees, as current screen readers' fall short of conveying information relating to spatial layout and helping users form a mental model of web pages congruent with that of their sighted partners. It is clear that the VI participants interviewed were fully aware of the drawbacks that the serial nature of screen readers impose on their web search activities. In fact, these challenges have led them to choose to perform some web search activities collaboratively when that was an option. In the interviews, sighted users tended to use more complex structures for storing retrieved information, such as under headings or multi-level lists, while VI users tended to use simpler flat or linear lists of information.

V. Implications

The studies we carried out highlighted the challenges encountered when VI and sighted users perform a collaborative web search activity. In this section we propose a number of design implications for the design of CCIS systems. Due to space limitations in this paper, we only present three.

1. Overview of Search Results

Developing a mechanism that provides VI group members with an overview of search results and the ability to focus on particular pieces of information of interest could help in increasing VI participants' independence in CCIS activities. VI web searchers are likely to perform the results exploration stage more effectively and efficiently if they could firstly get a gist of results retrieved and could then drill down for more details as required. This will advantage both individual and collaborative information seeking activities.

2. Cross-modal Shared workspace

Having a common place to save and review information retrieved can enhance both the awareness and the sense making processes and reduce the overhead of using multiple tools, especially in the case of VI users, who do not have sight of the whole screen at one time. The system should support a cross-modal representation of all changes made by collaborators in the shared workspace. As changes in a visual interface can be represent it in colours, changes in the audio interface might be represented by a non-speech sound or a modification to one or more properties of the speech sound, for example timbre or pitch.

3. Cross-modal representation of collaborators Search Query Terms and Search Results

Allowing collaborators to know their partner's query terms and results viewed will inform them about their partner's progress during a task. Additionally, having a view of partners search results can allow sighted users to collaborate with their VI partners while going through search results.

VI. Conclusion

This paper discussed CCIS; an area that has not previously been explored in research. The studies presented in this paper is a part of a project that aims to provide support to the CCIS process. The next part of the project is to investigate the validity of the design implications in supporting CCIS and their effect on collaborators performance and engagement.

References

[1] Morris, M. R. (2008). A survey of collaborative web search practices. In Proceedings of the twenty-sixth annual SIGCHI conference on Human factors in computing systems, New York, USA. ACM.

[2] Golovchinsky, G., Pickens, J., and Back, M. (2009). A taxonomy of collaboration in online information seeking. In JCDL Workshop on Collaborative Information Retrieval.

[3] Stockman, T., and Metatla, O. (2008). The influence of screen readers on web cognition. Proceedings of Accessible design in the digital world conference. York, United Kingdom.

[4] Sahib, N. G., Tombros, A., and Stockman, T. (2012). A comparative analysis of the information-seeking behavior of visually impaired and sighted searchers. Journal of the American Society for Information Science and Technology.

[5] Vigo, M., and Harper, S. (2013). Coping tactics employed by visually disabled users on the web. International Journal of Human-Computer Studies.

[6] Shah, C., and Marchionini, G. (2010). Awareness in collaborative information seeking. Journal of the American Society for Information Science and Technology.

Device-to-device (d2d) communication has emerged as a solution that promises high bit rates, low delay and low energy consumption which represents the key for novel technologies such as Google Glass, S Beam, and LTE-Direct. Such d2d communication has enabled computational offloading among collaborative mobile devices for a multitude of purposes such as reducing the overall energy, ensuring resource balancing across device, reducing the execution time, or simply executing applications whose computing requirements transcend what can be accomplished on a single device. While this novel computation platform has offered convenience and multiple other advantages, it obviously enables new security challenges and mobile network vulnerabilities. We anticipate challenging future security attacks resulting from the adoption of collaborative mobile edge cloud computing platforms, such as MDCs and FemtoClouds. In typical botnet attacks, “vertical communication” between a botmaster and infected bots, enables attacks that originate from outside the network. While intrusion detection systems typically analyze network traffic to detect anomalies, honeypots are used to attract and detect attackers, and firewalls are placed at the network periphery to filter undesired traffic. However, these traditional security measures are not as effective in protecting networks from insider attacks such as MobiBots, a mobile-to-mobile distributed botnet. This shortcoming is due to the mobility of bots and the distributed coordination that takes place in MobiBot attacks. In contrast to classical network attacks, these attacks are difficult to detect because MobiBots adopt “horizontal communication” that leverages frequent contacts amongst entities capable of exchanging data/code. In addition, this architecture does not provide any pre-established command and control channels (C&C) between a botmaster and its bots. Overall, such mobile device infections will circumvent classical security measures, ultimately enabling more sophisticated and dangerous attacks from within the network. We propose HoneyBot, a defense technique that detects, tracks, and isolates malicious device-to-device communication insider attacks. HoneyBots operate in three phases: detection, tracking, and isolation. In the detection phase, the HoneyBot operates in a vulnerable mode in order to detect lower layer and service-based malicious communication. We adopt a data driven approach, using real world indoor mobility traces, to evaluate the impact of the number of HoneyBots deployed and their placement on the detection delay performance. Our results show that utilizing only a few HoneyBot nodes helps detect malicious infection in no more than 15 minutes. Once the HoneyBot detects malicious communication, it initiates the tracking phase which consists of disseminating control messages to help “cure” the infected nodes and trace back the infection paths used by the malicious nodes. We show that HoneyBots are able to accurately track the source(s) of the attack in less than 20 minutes. Once the source(s) of the attack is/are identified, the HoneyBot activates the isolation phase that aims at locating the suspected node. We assume that the suspect node is not a cooperative device that aims at hiding its identity by ignoring all the Honeybot messages. Therefore, the HoneyBot requests wireless fingerprints from all nodes that have encountered this suspect nodes at a given time period. These fingerprints are used to locate these nodes and narrow down the suspect's location. To evaluate our localization accuracy, we first deploy an experimental testbed where we show that HoneyBots accurately localize the suspect node within 4 to 6 m². HoneyBots can operate efficiently in small numbers, as few as 2 or 3 nodes while improving the detection, tracking, and the isolation by a factor of 2 to 3. We also assess the scalability of HoneyBots using a large scale mobility trace with more than 500 nodes. We consider, in the attached Figure, a scenario of a corporate network consisting of 9 vulnerable devices labeled 1 to 9. Such network is attacked by one or many botmaster nodes using d2d MobiBot communication. We notice that attacks are propagated horizontally, bypassing all Firewall and intrusion detection techniques deployed by the corporate network administrators. In this scenario, we identify 3 main actors; the botmaster (red hexagon), the HoneyBot (green circle), the infected bot (red circle), and the cured or clear node (blue circle). We assume that the 9 nodes shown in the figure only represent the vulnerable d2d nodes in this corporate networks. We propose detection, tracking and isolation technique that aim at accurately and efficiently defend networks from insider d2d malicious communication.

Programming and coordinating decentralized ensembles of computing devices is extremely hard and error-prone. With cloud computing maturing and the emerging trend of embedding computation into mobile devices, the demand for building reliable distributed and decentralized systems is becoming increasingly common and complex. Because of these growing technical challenges, solutions for effective programming and coordination of decentralized ensembles remain elusive. Most main-stream programming methodologies only offer a node-centric view of programming, where a programmer specifies distributed computations from the perspective of each individual computing node (e.g., MPI, transactional memory, the Actor model, Linda Tuple Space, graph processing frameworks). When programming distributed computations in this style, programmers experience minimal shifts in paradigm but such concurrency primitives offer minimal support for the coordination problem. However, as systems grow in complexity and sophistication, maintaining code in this node-centric style of ten becomes costly, as the lack of concurrency abstraction means that programmers assumes all the responsibility of avoiding concurrency pitfalls (e.g., deadlocks and race-conditions). Because of this, ensemble-centric concurrency abstractions are now growing in popularity. In this style of programming, programmers are able to specify complex distributed computations from the perspective of entire collections of computing nodes as a whole (e.g., MapReduce, Google Web Tool-kit, choreographic programming), making implementations of distributed computations more concise and even making large classes of concurrency pitfalls syntactically impossible. However, programming distributed computations in this style typically require programmers to adopt a new perspective of computation. At times, they are overly restrictive and hence not applicable to a wider range of distributed coordination problems.

Our work centers on developing a concurrency abstraction to overcome the above challenges, by (1) providing a high-level ensemble-centric model of coordinating distributed computations, and (2) offering a clean and intuitive integration with traditional main-stream imperative programming languages. This framework as a whole, orthogonally combines a high-level concurrency abstraction together with established lower-level main-stream programming methodologies, maintaining a clean separation between the ensemble-centric concurrency model and the underlying sequential computation model, yet allowing them to interact with each other in a symbiotic manner. The benefit of this separation is twofold: first, a clear distinction of the elements of the coordination model from the computation model helps lower the learning curve of this new programming framework. Hence, developers familiar with the underlying main-stream computation model can incrementally build their technical understanding of the framework, by focusing solely on the coordination aspects of the framework. Second, by building the coordination model on top of an underlying main-stream computation model, we inherit all existing libraries, optimizations, as well as programming expertise available to it. We have addressed several key challenges of developing such an ensemble-centric concurrency model. In particular, we have developed a choreographic transformation scheme that transform our ensemble-centric programs into node-centric encodings, and we have also developed a compilation scheme that convert such node-centric encodings into lower-level imperative codes that can be executed by individual computing nodes. Finally, we proved the soundness of this choreographic compilation scheme by showing a two-step correspondence from ensemble-centric specifications to node-centric encodings, then to node-centric compilations. We have implemented an instance of this distributed programming framework for coordinating decentralized ensembles of Android mobile devices. This system is called CoMingle and it is built to integrate with Java and the Android SDK. The ensemble-centric nature of this programming abstraction simplifies the coordination of multiple Android devices, and in we demonstrate how the clean integration with Java and Android SDK allow local computations within each device to be implemented in a traditional manner, hence leveraging from an Android programmer's expertise, rather than forcing him/her to work with an entirely new programming environment. As proof-of-concept, we have developed a number of distributed mobile Android applications. CoMingle is open-source and available for download at https://github.com/sllam/comingle.