Arabic Similarity Search Tool for the OpenITI Corpus (SimilArITI) (v1-0)
Abstract
SimilArITI is a Python-based tool designed to identify textually similar passages across a corpus of Arabic works, mainly the OpenITI Corpus. It operates in two distinct phases: 1) an indexing phase, in which texts are preprocessed, segmented into fragments, and represented using TF-IDF vectorization at the fragment level; and 2) a search phase, in which a user-defined reference text is compared against the indexed fragments using cosine similarity.
Fields of Study & Subjects
Digital Humanities and Scholarship
Natural Language Processing Techniques
Topic Modeling
Publication details
- Journal / source
- DIGITAL.CSIC (Spanish National Research Council (CSIC))
- Publication date
- 27 January 2026
- Language
- EN