NTLM: a time-enhanced language model based ranking approach for web search

摘要: Time plays important roles in Web search, because most pages contain time information and a lot of queries are time-related. However, traditional search engines have little consideration on the pages. In particular, they do not take into account when ranking results. this paper, we present NTLM, new time-enhanced language model based algorithm for search. First, an effective to extract 〈keyword, content time〉 pairs pages, which associate each keyword page with appropriate time. Then introduce concept temporal tf, time-constrained term frequency, keyword. After that, propose measure similarity between temporal-textual basis combination textual relevance relevance. We conduct comparison experiments NTLM five competitor algorithms use two datasets, different types queries, metrics as MRR NDCG evaluate performance. The experimental results show that step extracting pairs, reaches high precision 93.2%, step, wins best respect NDCG.