A semidiscrete matrix decomposition for latent semantic indexing information retrieval
ACM Transactions on Information Systems (TOIS)
Exploiting hierarchical domain structure to compute similarity
ACM Transactions on Information Systems (TOIS)
Algorithmic detection of semantic similarity
WWW '05 Proceedings of the 14th international conference on World Wide Web
An architecture for personal semantic web information retrieval system
WWW '05 Special interest tracks and posters of the 14th international conference on World Wide Web
Generalizing PageRank: damping functions for link-based ranking algorithms
SIGIR '06 Proceedings of the 29th annual international ACM SIGIR conference on Research and development in information retrieval
| Hi-index | 0.00 |
Semantic retrieval for Web pages is an important issue in WWW environments. One of the typical methods for implementing semantic retrieval is to map Web pages onto an orthogonal space according to contents of Web pages. In this paper, we represent a space creation method for semantic retrieval for WWW environments. In this method, the orthogonal space, which is referred to as the retrieval space, is created by using sample Web pages. A sample Web page is selected from a Web page cluster as the typical Web page to stand all the Web pages in the cluster. As Web pages change rapidly every day, that is, new Web pages appear and some existing Web pages disappear, the sample Web pages have to be reselected dynamically according to the changes of Web pages. Following the changes of sample Web pages, the retrieval space has to be reconstructed dynamically. In general, the retrieval space reconstruction is a time consuming process because all the retrieval candidates in the space have to be re-mapped to the reconstructed retrieval space. In this paper, we present a mechanism for reducing the processing overhead of the retrieval space reconstruction. In our mechanism, we only reconstruct some parts of the retrieval space which are related to the modified sample Web pages. Furthermore, we present a fast search method to search Web pages which are needed to be remapped onto the retrieval space. By using these methods, our system is efficient for the semantic retrieval for Web pages in the WWW environment.