Relevance Feedback for Text Retrieval
Relevance feedback (RF) is a process by which the system, having retrieved some documents in response to the user’s query, asks the user to assess their relevance to his/her information need. The user’s relevance judgments are then used to either adjust the weights of the query terms, or add new terms to the query (query expansion).
Searchers may have difficulties in finding the words and phrases (terms) to express their information needs accurately and completely. They may also use different words in the queries than the words used by the authors of documents. On the other hand, searchers tend to know relevant information when they see it. In other words, it may be easier for them to tell which documents are relevant, instead of formulating a detailed query.
A typical relevance feedback process consists of the following steps: the user formulates and submits an initial query to an information retrieval system, which retrieves a ranked list of...
- 5.White RW, Ruthven I, Jose JM. A study of factors affecting the utility of implicit relevance feedback. In: Proceedings of the 31st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval; 2005. p. 35–42.Google Scholar