Particle Swarm Optimization for Punjabi Text Summarization

Abstract

Particle swarm optimization (PSO) algorithm is proposed to deal with text summarization for the Punjabi language. PSO is based on intelligence that predicts among a given set of solutions which is the best solution. The search is carried out by extremely high-speed particles. It updates particle position and velocity at the end of iteration so that during the development of generations, the personal best solution and global best solution are updated. Calculation within PSO is performed using fitness function which looks into various statistical and linguistic features of the Punjabi datasets. Two Punjabi datasets—monolingual Punjabi corpus from Indian Languages Corpora Initiative Phase-II and Punjabi-Hindi parallel corpus—are considered. The parallel corpus comprises 1,000 Punjabi sentences from the tourism domain while monolingual corpus contains 30,000 Punjabi sentences of the general domain. ROUGE measures evaluate summary where the highest measure, ROUGE-1, is achieved for parallel corpus with precision, recall, and F-measure as 0.7836, 0.7957, and 0.7896, respectively.

Automatic Text Summarization for Indian Languages

“…In 2021, Arti et al [6] proposed particle SWARN optimization for Punjabi text summarization. Using the proposed method, they achieved 79.59 ROUGE-1 score.…”

Automatic Text Summarization for Hindi Using Real Coded Genetic Algorithm

“…system summary ∩ re f erence summary re f erence summary (19) As stated in Equations ( 18) and (19), system summary ∩ reference summary stands for the number of overlapping words between the system summary and reference summary [71]. An F-measure metric is also called an F-score which is defined in ROUGE as the harmonic mean of the precision and recall as is given in Equation (20).…”

“…Statistical methods: Statistical-based text summarization [18][19][20] relies on the statistical distribution of specified features without understanding of the whole document.…”

A Computational Comparison of Three Nature-Inspired, Population-Based Metaheuristic Algorithms for Modelling-to-Generate Alternatives

“…Numerous metaheuristic approaches have been developed for use in a variety of decision-making environments (for some recent examples, see: Gergin et al 2019;Jain & Yada 2021;Murali et al 2022;Vasant et al 2020). For calculation and optimization purposes, Yang (2009Yang ( , 2010) created three population-based metaheuristics: the Firefly Algorithm (FA), the Bat Algorithm (BA), and the Cuckoo Algorithm (CA).…”