Find me opinion sources in blogosphere: a unified framework for opinionated blog feed retrieval

This paper aims to find blog feeds having a principal inclination towards making opinionated comments on the given topic, so that we can subscribe to them to track influential and interesting opinions in the blogosphere. One major challenge is assigning topic-related opinion scores to blog feeds, which is embodied in two aspects. Firstly, we should identify whether the blog feed has a principal on-topic opinionated inclination. This inclination should be collectively revealed by all posts of the feed. We should fully consider evidences from all the posts of the feed to identify salient information among many posts of the feed. Secondly, we should capture topic-related opinions in the blog feed while ignoring irrelevant opinions.

In this paper, we propose a unified framework for opinionated blog feed retrieval, which combines topic relevance and opinion scores with a generative model. Furthermore, we propose a language modeling approach to estimating opinion scores that is seamlessly integrated into the framework, where two language models, Topic-specific Opinion Model (TOM) and Topic-biased Feed Model (TFM), work collectively to reflect whether the blog feed shows a principal on-topic opinionated inclination. To estimate TFM, we propose a topic-biased random walk to exploit both content and structural information to capture topic-biased salient information in the feed. As for TOM estimation, we propose to use a generative mixture model with prior guidance to effectively capture topic-specific opinion expressing language usage. The conducted experiments in the context of the TREC 2009-2010 Blog Track show the effectiveness of our proposed approaches.