Information Retrieval Based on Formal Concept Analysis

摘要

In information retrieval, in order to improve the recall and precision, similar concepts are often processed during the information retrieval. With this regard, the possibility of evaluating concept similarity is acquiring an increasing relevance, since it allows the identification of different concepts that are semantically close. For every concept consist of two parts which include objects and attributes, so the computing model also has two parts. First part deals with the objects similarity by using hierarchy structure of concept lattice. The second part deals with the attributes similarity by using the expert's knowledge. Beside this, the depth also has relevance with concepts similarity. So the result is altered with concepts' depth. Experiment shows using concept lattice could improve information retrieval's recall and precision.