The Intelligent Retrieval System Based On Latent Semantic Analysis

Along with the rapid development of computer network technologies, large amount of information grows quickly on internet and brings much convenience for our life and business. Thus, it is getting more and more important for us to retrieve information accurately and efficiently today. Computers can send back satisfying information, unless they can understand our requirements accurately enough. On a certain extent, understanding user's requirements is to understand the words users input for searching. Therefore, it is necessary to develop an intelligent retrieval system based on semantic.This paper mainly researches on the theories and technologies of intelligent retrieval system and develops an intelligent retrieval system concerning geography knowledge based on Latent Semantic Analysis (LSA). This paper also improves on some algorithms of LSA: (1) the weight calculation algorithm is innovated to improve the information retrieval efficiency; (2) a K value selection automatic selection algorithm is put forward. (3) An approach search algorithm is put forward to make search results satisfying. (4) Due to the particularity of user's inputs, keywords factor is also taken into consideration. In addition, the main features of the intelligent retrieval system concerning geography knowledge are discussed in detail, including model design, development methods and key technologies. Finally, the experiment results of intelligent retrieval system are outlined. The system utilizes natural language processing technologies and realizes information retrieval at semantic level. Compared with traditional retrieval engines which mainly use keywords matching methods, this system shows its advantages in experiments.