In-depth blog retrieval based on improved pivoted normalization weighting coefficient
Jingyi Guan, Li Si, Weiran Xu, Guang Chen, Jun Hai Guo · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2012
The traditional blog retrieval which only focus on topical relevance is no longer satisfied by more and more blog users. They might be interested to follow bloggers whose posts express in-depth thoughts and analysis on the reported issues. This paper focuses on the problem of finding blogs that are relevant and in-depth about a user's query. We use LQtf coefficient, which is a kind of pivoted normalization weighting coefficient, to analyze the posts in blogs. And the effect of different kinds of in-depth analysis coefficient based on L-Qtf coefficient is also discussed. We propose an improved framework of in-depth facet blog distillation system in order to obtain in-depth blog to the query and set up experiments for comparison. Experimental results on the BLOG08 dataset show that the improved system is more effective than the prior system in TREC 2009 blog track.© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.