martinreynaert's repositories
HitPaRank
HitPaRank was developed as a tool to help select text units towards building a ground truth for concept-modeling which is domain expert controlled and geared towards replicability. It uses domain expert built lists of terms relevant to particular research questions applicable to a specific text corpus to help locate, extract and rank paragraphs from the corpus. It also helps defining and refining the actual term lists, adapting the theoretically relevant terms to the actual real-world forms as present in the corpus. It is currently geared towards working on corpora available in FoLiA XML format.