Title
kNN at TREC-9
Abstract
Abstract: We applied a multi-class k-nearest-neighborbased text classification algorithm to the adaptiveand batch filtering problems in the TREC9filtering track. While our systems performedwell in the batch filtering tasks, they did notperform as well in the adaptive filtering tasks,in part because we did not have an adequatemechanism for taking advantage of the relevancefeedback information provided by the filteringtasks. Since TREC-9, we have made considerableimprovements in our...
Year
Venue
Keywords
2000
TREC
adaptive filter
Field
DocType
Citations 
Information retrieval,Computer science
Conference
5
PageRank 
References 
Authors
0.50
2
2
Name
Order
Citations
PageRank
Tom Ault115619.83
Yiming Yang23299344.91