Title: Efficient data retrieval model based on semantic similarity analysis using chiroptera buzzard optimisation tuned deep CNN

Authors: Ankush Raosaheb Deshmukh; Premchand B. Ambhore

Addresses: Computer Science and Engineering, Government College of Engineering, VMV Road, Kathora Naka Amravati, Maharashtra, 444604, India ' Information Technology, Government College of Engineering, Amravati VMV Road, Kathora Naka Amravati, Maharashtra, 444604, India

Abstract: To extract meaningful insights, integrating the data classification and semantic text summarisation is essential, aiding in the identification of contextually significant content. Most of the existing techniques encounter multiple challenges from the perspective of machine understanding, especially for languages with limited resources, and fail to learn the sequence of correlations effectively. Nevertheless, there is still much space for enhancing the speed of data retrieved because current approaches fail to take the spatial and semantic aspects into account. To tackle this issue, this research presents an efficient data retrieval model utilising chiroptera buzzard optimisation adapted deep convolutional neural network (CBO adapted deep CNN) for semantic similarity analysis. Specifically, the chiroptera buzzard optimisation is utilised for feature selection and fine-tuning the hyperparameters of DCNN that improves the classification accuracy. Hence, the proposed model reduces the computational complexity and provides remarkable performance in terms of metrics attaining 99.98% accuracy, 99.53% recall, 99.93% precision, 99.84% Fbeta, 99.52% Cohen kappa, and 99.52% F1-score for 90% of training.

Keywords: data retrieval model; convolutional neural network; CNN; chiroptera buzzard optimisation; CBO; semantic similarity analysis; text summarisation.

DOI: 10.1504/IJBIC.2026.151782

International Journal of Bio-Inspired Computation, 2026 Vol.27 No.1, pp.17 - 30

Received: 19 Dec 2023
Accepted: 23 Jan 2025
Published online: 19 Feb 2026 *

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