Probabilistic information retrieval as a combination of abstraction, inductive learning, and probabilistic assumptions
- 2 January 1994
- journal article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Information Systems
- Vol. 12 (1) , 92-115
- https://doi.org/10.1145/174608.174612
Abstract
We show that former approaches in probabilistic information retrieval are based on one or two of the three concepts abstraction, inductive learning , and probabilistic assumptions , and we propose a new approach which combines all three concepts. This approach is illustrated for the case of indexing with a controlled vocabulary. For this purpose, we describe a new probabilistic model first, which is then combined with logistic regression, thus yielding a generalization of the original model. Experimental results for the pure theoretical model as well as for heuristic variants are given. Furthermore, linear and logistic regression are compared.Keywords
This publication has 8 references indexed in Scilit:
- A probabilistic learning approach for document indexingACM Transactions on Information Systems, 1991
- Models for retrieval with probabilistic indexingInformation Processing & Management, 1989
- Optimum polynomial retrieval functions based on the probability ranking principleACM Transactions on Information Systems, 1989
- A framework for effective retrievalACM Transactions on Database Systems, 1989
- PROBABILISTIC AUTOMATIC INDEXING BY LEARNING FROM HUMAN INDEXERSJournal of Documentation, 1984
- Relevance weighting of search termsJournal of the American Society for Information Science, 1976
- Precision Weighting—An Effective Automatic Indexing MethodJournal of the ACM, 1976
- On Relevance, Probabilistic Indexing and Information RetrievalJournal of the ACM, 1960