Hapax

Hapax analyses the vocabulary of software systems. I apply search engine technology (Latent Semantic Analysis) to analyse the vocabulary and topics of software. In my research, I used Hapax for

• aspect mining,
• software clustering and,
• source code classification. Hapax's approach is programming language independent as it is based on identifier names and comments only.  

---

Hapax Download

The original Hapax is written in Visualworks Smalltalk. Recently, Romain an me started porting Hapax to Squeak, but its not yet done, see

• http://www.squeaksource.com/Hapax.html

References

• Adrian Kuhn, Stéphane Ducasse and Tudor Gîrba, "Semantic Clustering: Identifying Topics in Source Code," Information and Software Technology (IST), vol. 49, no. 3, March 2007, pp. 230—243. (view Bibtex, download PDF)

• This journal paper is an extension of an WCRE paper and my Master's thesis.

Slides

• 2006-StuTS-LinguisticSoftwareAnalysis.pdf (best)
• 2005-WCRE-SemanticClustering.pdf