Companion page of "Querying Source Code with Natural Language" (Automated Software Engineering 2011)

This is the companion page of the paper "Querying Source Code with Natural Language". @inproceedings{Kimmig2011, title = {Querying Source Code with Natural Language}, author = {Kimmig, Markus and Monperrus, Martin and Mezini, Mira}, url = {https://hal.inria.fr/hal-00640496/file/paper.pdf}, booktitle = {{26th IEEE/ACM International Conference On Automated Software Engineering}}, pages = {376-379}, year = {2011}, doi = {10.1109/ASE.2011.6100076}, } Abstract One common task of developing or maintaining software is searching the source code for information like specific method calls or write accesses to certain fields. This kind of information is required to correctly implement new features and to solve bugs. For this, development environments offer search tool that are based on many radio-buttons and check-boxes. This paper presents an approach for querying source code with natural language. It enables the developer to execute a huge range of precise searches while being as easy and intuitive to use as writing natural language. The evaluation shows that the prototype implementation, integrated with the Eclipse development environment for Java, supports a wide range of queries and is able to correctly understand most real developer queries. Replication data Prototype: training-data.txt: Training data (annotated queries) mapping-data.txt: Mapping data (stemmed form -> JDT Search parameter) unit-tests.csv: Unit test queries Experiment: experiment-instructions.pdf: Instructions experiment-tasks.pdf: Task Description experiment-project.zip: Input Eclipse Project (Space invader) experiment-logs.zip: Logged input querie