This is a project that is still in progress.
Make a recommendation system for personalized promotional information.
#####1.The core of the reptile part.
#####2.The website homepage and comments are crawled.√
#####3.Use jdk's httpServer to provide services.√
#####4.~~Create a web-side project and collect user logs. ~~ (Inconvenient use)
#####5.Use weka for subsequent data processing. √
The finished csv dataset is in the data directory. #####Classification According to the view buy normal dislike classification, the data set is not balanced, using the smooth or cost-sensitive method does not feel very good; Temporarily according to the view buy dislike classification. #####result The best result of i148 is based on the decision tree of attribute selection, the correct rate is 82% The best result of i244 is Naive Bayes, with a correct rate of 77%.
#####1. Added b-crawl function, including: archive reply user fav #####2. Provide mxbean calls
For My Latest Project Visit https://www.reviewsoffers.com/