Search term relevance weighting- some recent results
Karen Spärck Jones
Abstract
Karen Spärck Jones
Abstract
This paper describes a series of experiments designed to test the value of search term relevance weighting schemes in the conditions typical of operational retrieval systems. The results show that relevance weighting can improve retrieval performance substantially in the stringent conditions implied by large document sets and little relevance assessment information.
OpenAlex reports 14 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper describes a series of experiments designed to test the value of search term relevance weighting schemes in the conditions typical of operational retrieval systems. The results show that relevance weighting can improve retrieval performance substantially in the stringent conditions implied by large document sets and little relevance assessment information.
Key concepts: Weighting, Relevance (law), Term (time), Computer science, Information retrieval, Value (mathematics), Relevance feedback, Data mining