The Analysis And Implementation Of Distributed Information Retrieval Engine

The purpose of establishing network is sharing data resource. At present, there are a lot of resources sharing server in Internet. Some resources are stored in each server for users'convenient access. These resources can be retrieved, and they can be accessed by users and other servers through web service. But existing problems, such as server's wide distribution and large number, it is difficult for users to retrieve the necessary information. In order to solve the problem, the distributed resources retrieval strategy is put forward. The main theme of distributed resources retrieval strategy is as follows:When the users retrieve information with any server, not only does the local resource can be retrieved in servers, but also the servers can be used as clients. One server, which is automatically to connect to other servers, will send retrieval request and combine the retrieval results from servers, then return the results to the users in the client.Based on the theory above, a distributed information retrieval engine is proposed in the paper. The network environment the engine used can be a LAN. In the network environment, each machine can be a Web server. A lot of resources are stored in the root directory of servers in the format of XML.When the user login any server in the network, and send out the request of retrieval, resource information, which is stored in an XML format, can be retrieved in the server and all servers connecting to the former server. The retrieval system has the advanced search function except general search function. The relevant resources can be retrieved according to the request of users. For example, the user can specify the type of resources to search and upload when advanced search function is used. In addition, display of search results in the system is further designed. and the resources can be previewed in the way of Word and PPT document. Also audio and video resources can be online played. Moreover, not only can the resources be downloaded by clicking the name of resources, but also some comments about them can be given. In order to increase the safety of the search system, user registration function is set in the system, it is convenient for both management and maintain of system. After many experiments, it is found that the information retrieval engine based on distributed technique in the paper has high performance of the retrieval and precision.