Proposal of privacy protection system for web forms using Bloom filter
Junpei Yamaguchi, Ryuya Uda · 2012
In recent years, the society has been built where even children can connect to Internet in their home, because of decrease in price of computers and of maintenance of communication infrastructure. On the other side, outflow of personal information is serious issue. Of course, there have already been some kinds of technology to analysis whether personal information is mentioned in user's diary on social network services such as MIXI, Twitter and so on. However, there are two problems; one is spending so much time examining the problem because of using Wikipedia as source of data base. The other is risk of server management against from peeking data because the method examines information which has already been sent to a server. Usui et al [9] suggested that browser plug-in method toward HTML form, used for BBS (Bulletin Board System), diary on social networking service and input form on Internet shopping site, taking advantage of morphological analysis and database searching that can find information which seems to be personal one. Using this method avoids sending information which includes privacy. Introduction of the method costs users little. However, the system requires too much time in making inquiries about database. Moreover word rating in scoring does not have a close relation to users. We would like to propose a system that has high-speed database with Bloom filter and the evaluation method that values relation between a user and location information of the user.