Search and evaluation methods for class level information retrieval
Andreas Henrich, Markus Wegmann · 2021
The task of expertise retrieval can be extended and generalized to a class search over collections; in this paper we describe the search for classes in the form of journals over a huge collection of articles. Given a textual query we aim to find thematically related and most appropriate classes (i.e. journals) for this request. To gain these suitable journals we utilize existing, new, and varied aggregation techniques introduced in expert search in which every article acts as a potential voter for its class - i.e. the journal it is published in.