Search

Recently, recommendation systems based on bipartite graph algorithm have been widely applied to many areas including E-business, but the weight of edge is ignored. Therefore, the commodity with high rating has not got the priority to be recommended. In order to solve the problem, we propose a personalized recommendation system based on user's interest. The results of web log mining are introduced to weighted bipartite graph, greatly improving the practicability of the recommendation.

Contrast with current WLAN (wireless local area network) security mechanism worked in data link layer. IPSec worked in IP layer is applied in WLAN as a security technology which is used in wired network has his superiorities. But IPSec uses Internet key exchange protocol (IKE) as default that need to a large number of computations is not fit for mobile devices of WLAN. This paper puts forward that using the third party authentication of symmetric algorithms to replace the major mode of the IKE first phase and maintain IKE original process of second phase basically. Doing so not only satisfy the WLAN application background but also maintain the original basic functions of IKE. We rename the improvement IKE to WKE (wireless key exchange). Theory analysis and experiment represent WKE has low calculation. It can improve the execution efficiency of the identity authentication and key negotiation between wireless station and IPSec security gateway.