Automated Feature Identification in Web Applications
This addresses the challenge for product managers in software companies to systematically reduce low-value features, though it appears incremental as it builds on existing partial methods.
The paper tackles the problem of feature creep in web applications by proposing an automated approach and tool for feature identification, validated through a case study on YouTube, Google, and BBC, indicating good potential for automation.
Market-driven software intensive product development companies have been more and more experiencing the problem of feature expansion over time. Product managers face the challenge of identifying and locating the high value features in an application and weeding out the ones of low value from the next releases. Currently, there are few methods and tools that deal with feature identification and they address the problem only partially. Therefore, there is an urgent need of methods and tools that would enable systematic feature reduction to resolve issues resulting from feature creep. This paper presents an approach and an associated tool to automate feature identification for web applications. For empirical validation, a multiple case study was conducted using three well known web applications: Youtube, Google and BBC. The results indicate that there is a good potential for automating feature identification in web applications.