Applying Adaptive Hybrid Recommendation Technology for Searching Algorithm Learning Articles

Abstract

In this generation, technology is developed in tremendous speed. The Information on the Internet is increasing at a high speed each day. People get plenty of information via the search engine and spend much time to filter out insignificant information at the same time. Therefore, this research system can filter all kinds of articles to exclude advertisement, news from network bookstore, or insignificant information related to keyword. After filtering, this system will gather up all useful articles and provide users to review. This study proposes the hybrid recommended system with multi-adaptive recommendation to learners. Hybrid recommendation is divided into two ways to recommend “Content-Based Recommendation” and “Collaborative Filtering Recommendation” articles. First, content-based recommendation is based on Term Frequency- Inverse Document Frequency to estimate the characteristic values of articles. Then, we set the weight of difficulty of keywords. After that, people can decide the level of article in the beginning and use it for reference. Besides, collaborative filtering recommendation is applied based on user abilities estimated by IQ tests, quizzes, online tests, ability certificates, and other exams. When the result of two users is similar and one of them thinks it’s useful, the article will be automatically forwarded to the other.