Abstract
Latent Semantic Analysis (LSA) is a retrieval technology based on the latent concept indexing. The theory and method of LSA are summarized, and the underlying mathematical meaning is explained. The effect of LSA is then illustrated by the change of the correlation between documents or terms in a simple example. Moreover, in order to check the retrieval efficiency of LSA, the concept of self-indexing matrix is presented. The practicability of LSA in the field of information retrieval is finally indicated by the experimental results.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 107-111 |
| Number of pages | 5 |
| Journal | Huanan Ligong Daxue Xuebao/Journal of South China University of Technology (Natural Science) |
| Volume | 32 |
| Issue number | SUPPL. |
| State | Published - Nov 2004 |
| Externally published | Yes |
Keywords
- Chinese information processing
- Latent Semantic Analysis
- Latent semantic space
- Singular value decomposition
ASJC Scopus subject areas
- Engineering(all)