Backlog Composition in an Agile Context


Context

The main idea of agile-driven development (independently of methods such as SCRUM, Kanban, XP) is to continuously deliver value to the end-user. To perform such a task and specify the product to be developed, software developers (SDs) and product owners (POs) classically rely on the definition of user stories, organized in a product backlog.

In this project, we want to leverage backlogs (e.g., product backlog, sprint backlogs) to provide feedback to both SDs and POs for a given project. Natural language processing tools can be used to extract information from user stories (e.g., Visual Narrator). Based on the extracted data, it is possible to infer a domain model (similar to an UML class diagram) for each story. this domain can be exploited to give useful information to SDs and POs, for example:

The project follows the development of a proof of concept that binds together VisualNarrator (an NLP tool to analyze stories) and Neo4J (a graph database). We want to go deeper in the analysis of the backlogs, by digging into realistic ones and leverage POs knowledge.

Skills and Background

Required role of the student

References

Supervision

Sébastien Mosser
Professor, UQAM