Query Formulation for Prior Art Search - Georgetown University at CLEF-IP 2013

Our group participated in the CLEF-IP 2013 Passage Re- trieval starting from Claims task. We focus on formulating representative queries from various metadata that is embedded in a patent document. We then submit the queries to a state-of-the-art search engine to per- form document level retrieval. For passage level retrieval, we implement a TF-IDF algorithm that calculates the sum of query keywords' TF- IDF scores. We submitted six runs, which tested dierent uses of the metadata and dierent retrieval algorithms. We nd that carefully con- structed structured queries from titles and terms with mid-range IDF values are eective for patent prior art retrieval.