Title: Soft-computing method for settling land disputes cases based on text similarity

Authors: Okure U. Obot; Faith-Michael E. Uzoka; Anietie E. John; Samuel S. Udoh

Addresses: Department of Computer Science, University of Uyo, PMB 1017, Uyo, Nigeria ' Department of Mathematics and Computing, Mount Royal University, Calgary, Canada ' Department of Computer Science and Mathematics, Ritman University, Ikot Ekpene, Nigeria ' Department of Computer Science, University of Uyo, PMB 1017, Uyo, Nigeria

Abstract: Incessant delays in the administration of justice in land disputes cases caused by inability to access similar cases often lead people to take laws into their own hands, resulting in wanton destruction of lives and property. In this study, three similarity models, namely: cosine, Jaccard, and text semantic similarity (TSS) and fuzzy logic technology were reviewed and applied to 205 settled cases of land disputes collected from the high court of justice in Ikot Ekpene, Nigeria. Our study revealed that the cosine similarity measure had the strongest correlation, (72%) followed by Jaccard (70%), fuzzy logic (70%) and TSS (63%). Considering this, we recommend fuzzy logic combined with cosine for the building of a legal case-based reasoning system.

Keywords: text analysis; legal cases; cosine; Jaccard; text semantic similarity; TSS; fuzzy logic; natural language processing; NLP.

DOI: 10.1504/IJBIS.2023.132073

International Journal of Business Information Systems, 2023 Vol.43 No.3, pp.369 - 393

Received: 24 Mar 2020
Accepted: 22 Jul 2020
Published online: 11 Jul 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article