Guiding Software Developers by Social Networking Application Plug-in using the Multiple Bridge Source Repository through a Data Mining Integrated Approach

TL;DR

This paper introduces SNAP, a plugin system that helps social networking app developers by mining multiple repositories to recommend relevant, error-free API usage patterns, thereby simplifying app development.

Contribution

It proposes a novel framework that bridges multiple source repositories and applies data mining to assist developers with API pattern recommendations.

Findings

SNAP effectively recommends relevant API sequences.

The system provides error-free and understandable API patterns.

It supports reuse and modification of API sources.

Abstract

In today's world, social networking is an important (power full) medium of mass communication. People of almost all classes have been interacting with each other and sharing their views, moments, and ideas by using enormous user-friendly applications in different social networking sites. It's really unbelievable to find a person who never heard about the social network. The available social networking sites usually opportune their users to develop various customized applications through particular templates and embedded sources of codes. The users with average knowledge of development often encounter difficulties to reuse those resources and eventually lack guidelines and necessary API recommendations. In our work, we have proposed a framework and model to help those apps developers through a user assistance plug-in tool that is able to provide identical API usage patterns and sequences in response to a particular user query. We have titled our system as a Social Networking Application Plug-in (SNAP). We search social networking apps repository where multiple storages are bridged and apply respective mining algorithm to find the relevant sequences to fulfill the user needs. It provides similar, most relevant, and functional API usage scenarios as well as gives an option to choose, reuse, and modify the recommended sources. From investigations we have ever made, our SNAP approach is capable to recommend users' error-free, understandable, and minimal API patterns.