3XL News: A Cross-lingual News Aggregator and Reader

We present 3XL News, a multi-lingual news aggregation application for iPad that provides real-time, comprehensive, global and multilingual news coverage. Using methods, developed within the XLike project, for semantic data extraction from news articles an

  • PDF / 2,464,090 Bytes
  • 6 Pages / 439.37 x 666.142 pts Page_size
  • 48 Downloads / 177 Views

DOWNLOAD

REPORT


. We present 3XL News, a multi-lingual news aggregation application for iPad that provides real-time, comprehensive, global and multilingual news coverage. Using methods, developed within the XLike project, for semantic data extraction from news articles and linking of news stories we are able to construct a concise, yet in-depth view of current news stories and their semantic relation. This enables users real-time monitoring of current global events and analysis of diverse reporting in different languages and navigation across related news stories.

1

Introduction and Motivation

Real-time access to the latest news any time and from any location has become possible with widespread adaptation of mobile networks and devices. Increasingly, users of such devices expect custom-made, native applications to access their information. In this article we describe 3XL News: a system stemming from the XLike EU project, offering X(cross)Lingual analysis of eXtra Large News. 3XL News is an iOS application targeting news professionals and the general public. It shows how semantic technologies can be used in a real-world scenario to provide real-time global news monitoring and analysis across several languages. The main novelty is the linking of stories across six languages using semantic data derived from multi-lingual entity detection and cross-language news linking. Related Work: A multitude of news monitoring mobile applications is available on the market, broadly divided into two groups: publishers’ own applications (like BBC, Al-Jazeera and RTV Slovenija) and news aggregators (like News Republic, Yahoo News, EMM mobile app [7] and iDiversiNews [8]). Those services reduce the overwhelming amount of information by summarizing news events and filtering them according to predefined user preferences or reading habits, however, they do not link news stories across languages. There are systems that perform c Springer International Publishing Switzerland 2015  F. Gandon et al. (Eds.): ESWC 2015, LNCS 9341, pp. 3–8, 2015. DOI: 10.1007/978-3-319-25639-9 1

4

E. Belyaeva et al.

cross-language linking such as NewsReader [10] and SPIGA [4] but support only four and two languages respectively and lack the capability to explore news along other dimensions (sentiment, location etc.).

2

Technological Background

XLike Project. XLike stands for Cross-LIngual Knowledge Extraction. The main goal of the project1 was to develop technology to monitor and aggregate knowledge spread across mainstream and social media, as well as across different languages. This is achieved by applying computational linguistics and semantic technologies to extract formal knowledge from multilingual texts. Developed methods were integrated into an extensive linguistic and semantic text analysis pipeline [2], which provides input data for 3XL News. News is gathered from the Internet [9], annotated with semantic data [2], similar articles are linked across languages [5], news articles are clustered into stories and finally, news stories are linked across languages using