Abstract:
This research analyzes how selecting the most suitable open source software among a variety of identical products has been a great challenge due to the uneven quality and the rapid development. To overcome the issue this project, propose a deep learning approach focusing to assist the open source selection process. The proposed solution is a web-based solution presenting an aspect-based sentiment analysis framework which can suggest the best possible solution which fits the user requirement. This research consider software aspects including performance, usability, testability and user interface aesthetics. Multiple data sources are incorporated to obtain different view-points utilizing user feedback and enhance the reliability. The solution uses Deep learning, Natural Language Processing and Sentiment analysis to accomplish this task.
