Determining User Intent from Query Patterns
Topics: Ranking, Search Intent, Search Query Processing
The document published on October 21, 2014, with Yair Kurzion from Mountain View, California, listed as the inventor and Google Inc. as the assignee. The patent, titled “Determining User Intent from Query Patterns,” details methods, systems, and program products for identifying user intent based on query patterns. It involves analyzing current queries including a query term, obtaining search results, and associating an intent with the session based on the submission of a query containing a refining term. This approach includes identifying prior sessions with matching queries and refining terms, and modifying scores of documents in search results based on their frequency of selection in prior sessions compared to a general population. The document suggests a refined way of ranking search results by better understanding user intent, potentially enhancing the accuracy and relevance of search outcomes.