York_University at TRECVID 2010.
Jun Miao, Jimmy Xiangji Huang, Qinmin Hu · 2010
In this paper, we describe our work done by members at York University in Canada for the KIS (Known-item search) task of TRECVID 2010. This is the first time that we participate in the TRECVID. With rich experience in text retrieval, we mainly focus on the meta information of videos, and try to figure out the importance of these description corpus. In order to obtain this goal, we do not use any video or audio technologies. Only text retrieval methods are utilized to find the know items. Traditional weighting models in text retrieval like BM25 and Lemur TF-IDF are used. Meanwhile, we also use query expansion methods to improve the performance. However, the results are not promising. We make a further discussion about the reason at the end of this paper.