Using context to improve search precision

Search techniques are methods used for retrieving information that a user is looking for. It is vital that these techniques are efficient and help users locate the information that they need. Current search techniques do not return results that are precise and contextually appropriate for the user's environment. This thesis describes a search technique that utilizes appropriate contexts in answering user queries. Context information is used to filter out intermediate query results that are not relevant to the user, while preserving results that are accurate, thus improving the precision of the search results. A prototype system demonstrating this technique has been implemented using the Oracle Expression Filter to assist with context inference. A number of experiments were conducted to validate this search technique by measuring the precision increase of query results.