Analyzing Tag-Based Mashups with Fuzzy FCA

Mashing up with APIs has become a critical skill for Web 2.0 developers to construct creative web applications quickly. With the help from descriptive tags, developers can search the target Web APIs easier than before. However, it is still difficult to determine proper tags that best describe Web API. This paper is focusing on analyzing semantic tags and the relationships between each Web APIs. We apply Fuzzy Formal Concept Analysis (FFCA) and Bidirectional Co-occurrence approach to analyze tag relations automatically. As the experiment results shown, Bidirectional Co-occurrence enhanced FFCA model can increase the accuracy of tag mining, and FFCA could speed up the computation. Our Bidirectional Co-occurrence approach is more efficiency and accurate than native approaches.