Skip to content

Youxise/Document_Matching.IR

Repository files navigation

Document matching project

In this project, I developed an application using Tkinter for score matching between LISA collection documents by employing information retrieval techniques. The project includes several components: I focused on descriptor and inverse creation, added weight assignment, developed 3 models :

  • The vector space model (with 3 options : Scalar product, Cosine & Jaccard measures)
  • The probabilistic model (BM25)
  • The boolean model

Finally, I addressed metrics and the precision-recall curve implementation for the evaluation part.