Home

Welcome to the homepage of KaVE project. What started in the Software Campus at 2013 (a German research program that encourages intensive exchange between academia and industry) significantly grew in size over the years. The KaVE project has evolved into a stable platform, specifically for research in the area of "recommendation systems for software engineering", but also to answer more general questions about human factors in software engineering.

In which topic are you interested in?

MSR Mining Challenge

Our interaction dataset has been selected as the mining challenge for the "International Conference on Mining Software Repositories" (MSR'18).

FeedBaG++

A general purpose interaction tracker. Ready to be installed in Visual Studio to capture interactions of its user and easy to extend to capture additional context information during development.

CARET

Platform tailored to the needs of the development of recommendation systems for software engineering. Provides tooling and the means for a proper evaluation.

Intelligent Code Completion

Developers have to work with new APIs daily, this just got easier! Based on the powerful PBN engine, developer can enjoy intelligent API proposals in Visual Studio by installing the tool.

Datasets

Research on RSSE heavily relies on data. We provide a dataset of pre-analyzed Github repositories that can serve as reference data for training a model. In addition, we provide a dataset of developer interactions, which can be used for evaluation.

You will also find many other topics on this website. For example, we maintain a list of projects in which we applied our tools. You might an in depth discussion of our core data structures Simplified Syntax Trees and Enriched Event Streams helpful.