oa Relate-me: Making News More Relevant

To get readers of international news stories interested and engaged, it is important to show how a piece of far-away news relates to them and how it might affect their own country. As a step in this direction, we have developed a tool to automatically create textual relations between news articles and readers. To produce such connections, the first step is to detect the countries mentioned in the article. Many news sites, including Al Jazeera (http://www.aljazeera.com), use automated tools such as OpenCalais (http://opencalais.com) to detect place references in a news article and list those as a list of countries in a dedicated section at the bottom (say, List A: [Syria, Germany]). If not already included, relevant countries could be detected using existing tools and dictionaries. The second step is to use the reader's IP address to infer the country they are currently located in (say, Country B: Qatar). Knowing this country gives us a "bridge" to the reader as now we can try relate the countries from List A, to the reader's country, Country B. Finally, we have to decide which type of contextual bridges to build between the pairs of countries. Currently, we are focusing on four aspects: 1) Imports & Exports: this section displays imports and exports between the Country B and List A, if any. For instance: "Qatar exports products worth $272m every year to Germany, 0.27% of total exports." Upon clicking on this information, it will redirect the user to another website, showing a breakdown of these imports and exports. 2) Distances: this simply states the direct distance in kilometers from Country B to every country in List A. For instance: "Syria is 2,110km away from Qatar." Upon clicking on this information, it will navigate to a Google Maps display, showing this distance. 3) Relations: this provides a link to the designated Wikipedia page between Country B and every country in List A. For instance, "Germany - Qatar Relations (Wikipedia)." It also shows a link relating the countries using Google News: "More on Qatar - Germany (Google News)" 4) Currencies: this shows the currency conversions between Country B's currency and every other country in List A, for instance: "1 QAR = 0.21 EUR (Germany's currency)." Our current tool, which will be demonstrated live during the poster presentation, was built using JavaScript. With the use of tools such as Greasemonkey, this allowed us to test and display the results of the project on Al Jazeera (http://www.aljazeera.com), without having site ownership. We believe that the problem of making connections between countries explicit is of particular relevance to small countries such as Qatar. Whereas usually a user from, say, Switzerland, might not be interested in events in Qatar, showing information regarding trade between the two could change their mind.