Venkatesh presented the work at WSDM'17

Venkatesh presented his recently accepted work "ANNE: Improving Source Code Search Using Entity Retrieval Approach" at International Conference on Web Search and Data Mining, WSDM 2017. The conference was held in Cambridge, UK. The acceptance rate of the conference is 16%. 
Code searches with natural language (NL) terms, currently, perform poorly because programming concepts do not always lexically match their syntactic forms. In this work, we present a programming language agonistic technique that leverages the structural similarities in how people phrase (programming) questions and the repetition of syntactic structures in source code, to map source code lines to their programming concepts. We implement our approach in a tool called ANNE, that includes this mapping for C and Java. With the use of NL terms, participants obtain better completeness scores, and are significantly quicker, a 29% reduction, than in the control condition.
For more details, refer to the paper here.

Links to tools created by the group

Get in touch with us

Education - This is a contributing Drupal Theme
Design by WeebPal.