Collocation extraction is an elaborate problem in the eld of corpus linguistics that requires both statistical and linguistic information in order to be successful. The problem is attracting more attention as the importance of collocations, (aka multi-word expressions) is recognized by the NLP community. In this report, I present the work that I reviewed, and the Colloc module that was the result of this work, by explaining reasons behind taken decisions during the implementation, and of course, what remains to be done.