In concept lattice-based information retrieval, the task of document retrieval is being supported by means of Formal Concept Analysis. The steps of query matching and document ranking are informed by an underlying concept lattice thus induced from a set of documents. Recently, Codocedo et al. proposed a refined approach to this using semantic similarity as the basis for query evaluation and document ranking. This talk will give an introduction to their approach and discuss evaluation results. The talk closes with a current overview on related research done by Dominik, also pointing to future work.
02.02.2017 - 10:15