Abstract:This paper is to study the possibility and effectiveness to combine vector space model and probabilistic model so as to achieve the improved IR performance. Analyses are firstly made to observe the necessity and possibility of combining the two models. Two systems were used in the study: Okapi for probabilistic model and SMART for VSM. Evaluation on three IR standard test collections, namely CACM, ADI and MED, show that: (1) VSM based approach we observed always has 13% to 24% smaller on average rank than probabilistic based approach. (2) Probabilistic approach is consistently 6.6% to 44.2% lower on unfound proportion. (3) Results of 11-point average precision and top-5 precision do not have obvious difference between two systems. (4) For each test set, there are 8% to 20% relevant documents could be retrieved by only one system. These studies confirmed the necessity of our idea to combine the two models. The performance of the combination criteria were also evaluated on the three collections. Experimental results show that the combined criterion consistently leads to at most 6.9% improvement on 11-average precision and 9.2% improvement on top-5 precision compared with the better performed model.

An Improved VSM Based Information Retrieval System and Fuzzy Query Expansion

A Semantic Query Expansion-Based Patent Retrieval Approach

Exploiting Semantic Knowledge Base for Patent Retrieval

Merging case relations into VSM to improve information retrieval precision

Efficient information retrieval based on a combination of vector space and probabilistic models

Design and Application for the Model of Semantic Query Expansion Based on Domain Ontology.

A patent retrieval method based on automatic query expansion

A Revisit Of Query Expansion With Different Semantic Levels

Color Image Retrieval Utilizing Extended Fast Vq Codeword Search Technique And Vector Composition-Based Feedback

A Relevance Evaluation Model of Information Retrieval Based on Weighted Term Frequency

Research and Implementation of a Full-text Retrieval System

Research on Semantic Query Expansion with Temporal Information

Concept based query expansion using hidden markov model

Selecting Expansion Terms As a Set Via Integer Linear Programming

Semantic Query Expansion and Context-Based Discriminative Term Modeling for Spoken Document Retrieval

Concept Based Query Expansion Using Wordnet

VectorSearch: Enhancing Document Retrieval with Semantic Embeddings and Optimized Search

Concept Based Query and Document Expansion Using Hidden Markov Model

Query Expansion Based High Performance Chinese Voice Retrieval

Improving Pseudo-Relevance Feedback in Web Information Retrieval Using Web Page Segmentation.

Evaluate Retrieval Systems Based on Ontology Vocabulary.