Artificial Intelligence
Natural language processing
Semantic web

Thematic exploration of a corpus of Quebec articles in Humanities and Social Sciences

Student: Arthur Tobler

Supervisor: Catherine Beaudry

Co-supervisor(s): Michel Gagnon

This work aims at two objectives. First, it provides a thematic representation of a corpus of SHS articles using unsupervised semantic extraction models. Such representations can be used to carry out an efficient document search within the collection. Then, we seek to describe the thematic evolution within this corpus, i.e. how research ideas and interests change over time. This temporal description can help to characterize the influence of an article in a field and thus provide a new approach to measuring the impact of an author, complementary to bibliometric approaches using citation analysis.