Single document extractive text summarization using cuckoo search algorithm

Pati, Siba Prasad; Rautray, Rasmita

doi:10.1080/02522667.2022.2094547

Cited by 1 publication

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Pati and Rautrayet al, 2022 [33] have shown a comparative study between the Cuckoo search algorithm (CS), Firefly algorithm (FF) & Flower pollination algorithm (FP) for reducing the contents of the original document using the DUC 2003 dataset. This research was meant for single document extractive type summarization.…”

Section: Selvanmentioning

confidence: 99%

Single Document Text Summarization Based on the Modified Cat Swarm Optimization (Mcso) Algorithm

Rautaray,

Panigrahi,

Nayak

2024

Preprint

View full text Add to dashboard Cite

In today's digital age, the World Wide Web has provided us with an immense volume of online information. Consequently, the challenge of extracting relevant insights from this vast data has emerged. Recently, text summarization has gained recognition as a solution for distilling valuable orderas of extensive credentials. Depending on the number of credentialsmeasured in favor of summarization is partitioned as single-document and multi-document summarization, which is a complex challenge for researchers in achieving accurate summaries. This research methodology, considered single document summarization by using the following steps they are text-pre-processing, Feature extraction, vectorization, and Modified Cat Swarm Optimization (MCSO) algorithm. Initially, the input documents are pre-processed for cleaning the data and they contain sentence segmentation, word tokenization, stop word removal, and lemmatization. In feature extraction, a score of the sentence is computed using Time Frequency-Inverse Domain Frequency (TF-IDF) also then words are formed toward a vector. After that, the generated vectors are post-processed, and Finally, the Modified Cat Swarm Optimization (MCSO) algorithm is utilized toprovideasummary of the single document with its fitness function. The performance of the model was compared with existing approaches such as Spherical, Rastrigin, and Griewank. To assess their effectiveness, we use benchmark datasets of the Document Understanding Conference (DUC) and evaluate algorithms using diverse metrics such as ROUGE score, F score, and summary accuracy, as well as best-case, worst-case, and average-case analyses. The experimental analysis demonstrates that Rosenbork surpasses other summarization methods included in the study.

show abstract

Section: Selvanmentioning

confidence: 99%