A new method for information retrieval, based on the theory of relative concentration
SIGIR '90 Proceedings of the 13th annual international ACM SIGIR conference on Research and development in information retrieval
Shortest-substring retrieval and ranking
ACM Transactions on Information Systems (TOIS)
| Hi-index | 0.00 |
We show that the similarity measures for p-norm retrieval, as defined by Salton, Fox and Wu have some undesirable mathematical properties. We propose a new function that remedies some of these drawbacks. Still, even for this new similarity measure the extended Boolean model has some properties which can only be described as 'heuristic'.