In this paper we describe the methods that we have developed for the technology survey and prior art search tasks of TREC
2009 Chemical IR Track. We studied various approaches for both tasks. In particular, for the technology survey tasks, we
utilized structured retrieval, query expansion with synonyms of the detected chemical entities, rule based entity detection and
filtering techniques. For prior art search task, we used feature selection (to select the query terms for transforming a query
patent into a search query), structured retrieval, query expansion with synonyms of the detected chemical entities, date
filtering and re-ranking of the results by utilizing IPC code similarities. Both of our approaches have an acceptable
performance but still leave room for improvement.