This paper addresses blog feed retrieval where the goal is to retrieve the most relevant blog feeds for a given user query. Since the retrieval unit is a blog, as a collection of posts, performing relevance feedback techniques and selecting the most appropriate documents for query expansion becomes challenging. By assuming time as an effective parameter on the blog posts content, we propose a time-based query expansion method. In this method, we select terms for expansion using most relevant days for the query, as opposed to most relevant documents. This provide us with more trustable terms for expansion. Our preliminary experiments on Blog08 collection shows that this method can outperform state of the art relevance feedback methods in blog retrieval.
[1]
Jaime G. Carbonell,et al.
Retrieval and feedback models for blog feed search
,
2008,
SIGIR '08.
[2]
W. Bruce Croft,et al.
Relevance-Based Language Models
,
2001,
SIGIR '01.
[3]
Jong-Hyeok Lee,et al.
An improved feedback approach using relevant local posts for blog feed retrieval
,
2009,
CIKM.
[4]
ChengXiang Zhai,et al.
Positional language models for information retrieval
,
2009,
SIGIR.
[5]
Craig MacDonald,et al.
Overview of the TREC 2009 Blog Track
,
2009,
TREC.
[6]
Maarten de Rijke,et al.
Bloggers as experts: feed distillation using expert retrieval models
,
2008,
SIGIR '08.