BJUT at TREC 2015 Microblog Track: Real-Time Filtering Using Non-negative Matrix Factorization

In this paper, we described our approaches to the RealTime Filtering Task in the TREC 2015 Microblog track. We submitted the results for scenario B: periodic email digest. In this ad hoc search task, we introduced a real-time filtering framework using non-negative matrix factorization. To build this framework, we firstly considered the Real-Time Filtering Task as a short text retrieval problem, and proposed a non-negative matrix factorization based Microblog retrieval model (NMF Framework). Then after a review of the potential influencing factors in Microblog retrieval, the main influencing factor, i.e., short query expansion, was modeled as the additional regularized constraint items in NMF Framework. Experimental results show the proposed approach is better than classical method in microblog real-time filtering with the above-mentioned additional regularized constraint items.