Creating a reliability scoring extension for the MONAx event extractor
|has title::An algorithm for calculating the reliability of news events on the web|
|Master:||project within::Information Sciences|
|Student name:||student name::Marc Jacobs|
|Second supervisor:||Thomas Ploeger|
|Second reader:||has second reader::Davide Ceolin|
|Company:||has company::Vrije Universiteit|
The World Wide Web offers us a large amount of unstructured news. Several attempts are made to provide more structure and semantics to this unstructured pool of news, for example projects like MONAx and Newsreader. These projects primarily focus on the extraction of entities from the source text and adding annotations to them. Although this is a great first attempt in structuring news articles on the web, other challenges arise. For instance, little research is done in the field of the reliability of the news sources.
The ease of publishing news through media like blogs and twitter makes everyone a potential journalist. Except, just a small amount of these publishers are actual journalists. Therefore, biased and unreliable news is becoming a big problem. In this research we try to find a "reliability algorithm" used to calculate a reliability score related to a news event. Therefore, the following main research question is formulated: How can the reliability of online news articles be measured, with activist news events in particular?
The sub questions for related to the question above are: What factors influence the reliability of news sources on the web? How will the reliability results be stored in the data model? How can the reliability results be aggregated to indicate the reliability of an event?
Research areas: Semantic web, Information Retrieval, Data Visualization, Social Sciences