5th International Workshop on News Recommendation and Analytics (INRA 2017) will be held in conjunction with IEEE/WIC/ACM International Conference on Web Intelligence (WI), 23-26 August 2017, Leipzig, Germany.
Submission deadline extended: 21 May 2017
As the amount of data on the internet increases it is getting harder to find the information that people are looking for. Recommender systems are built to bring the most relevant information to users within the huge amount of data on the internet using the users’ personal interests and preferences. Even though there is steady progress in recommender systems and also visible progress in news recommender systems, there are many challenges that need to be solved or improved for the systems to receive widespread acceptance. Compared to recommender systems in domains like music, movies and books, news recommender systems pose some particular challenges that call for new and deeper analyses of both users and content: The news domain is marked by (i) dynamic streams of news articles where different news sources on the internet publish hundreds of new articles every hour, (ii) willingness to read news articles that are independent from user interests like breaking news, (iii) unstable user interests that change much faster than in other domains (the taste of movies or food of a user takes years to change), (iv) recency issues that render old news stories less interesting than recent ones, and (v) unstructured subjective content that create content analysis problems and may turn recommendations unreliable. These issues also complicate the modeling and monitoring of user interests and preferences, since users are not giving explicit signals of their interests and information about users need to be deduced from their observed attitude towards news. The news domains intrinsic complexities combined with the commercial interests of media companies is a good basis for innovative approaches to both news content analysis and news recommendation.
The news domain is characterized by a constant flow of unstructured, fragmentary, and unreliable news stories from numerous sources and different perspectives. Finding the right information, either in terms of individual news stories or aggregated knowledge from analyzing entire news streams, is a tremendous challenge that necessitates a wide range of technologies and a deep understanding of user preferences, news contents, and their relationships.
This workshop addresses primarily news recommender systems and news analytics, with a particular focus on user profiling and techniques for dealing with and extracting knowledge from large-scale news streams. The news streams may originate in large media companies, but may also come from social sites, where user models are needed to decide how user-generated content is to be taken into account. As part of news recommendation and analytics, Big Data architectures and large-scale statistical and linguistic techniques are used to extract aggregated knowledge from large news streams and prepare for personalized access to news. Topics of interests include but are not limited to:
News semantics and ontologies
News summarization, classification and sentiment analysis
Recommender systems and news personalization
Group recommendation for news
User profiling and news context modeling
Real time news recommendations
News evolution and trends
Large-scale news mining and analytics
Evaluation methods
News from social media
Big Data technologies for news streams
News recommendation and analytics on mobile platforms
Privacy in news recommender systems
Gamification in news recommender systems
News diversity
Detection and analysis of fake and/or trusted news