Generative models such as statistical language modeling have been widely studied in the task of expert search to model the relationship between experts and their expertise indicated in supporting documents. On the other hand, discriminative models have received little attention in expert search research, although they have been shown to outperform generative models in many other information retrieval and machine learning applications. In this paper, we propose a principled relevance-based discriminative learning framework for expert search and derive specific discriminative models from the framework. Compared with the state-of-the-art language models for expert search, the proposed research can naturally integrate various document evidence and document-candidate associations into a single model without extra modeling assumptions or effort. An extensive set of experiments have been conducted on two TREC Enterprise track corpora (i.e., W3C and CERC) to demonstrate the effectiveness and robustness of the proposed framework.
展开▼
