Current search engines generally employ link analysis techniques to web-page re-ranking. However, the same techniques are problematic in small webs, such as websites or intranet webs, because the link structure is different from the global Web. We propose a re-ranking mechanism based on implicit links generated from user access patterns. Experimental results indicate that our method outperforms keyword-based method, explicit link-based PageRank and DirectHit. Categories and Subject Descriptors H.3.3 [Information Storage and Retrieval]: Information Search and Retrieval - search process, retrieval models; H.2.8 [Database Management]: Database Applications - Data mining General Terms Algorithms, Experimentation Keywords Small web search, user access pattern, link analysis, PageRank.