Title: Using free open-source tools for text visualisation over unstructured corpus effectively

Authors: Gowri R. Choudhary; Iti Sharma

Addresses: Computer Science Department, Career Point University, National Highway 52, Opp Alaniya Mata Ji Mandir, Alaniya, Kota, Rajasthan 325003, India ' Computer Science Department, Government Polytechnic College, Kota, Rajasthan, India

Abstract: Text visualisation is an essential analytical task in many applications. Researchers from such fields can utilise the graphical output but need tools for its generation due to lack of expertise. Though several open-source tools are available, there is a challenge in choosing a suitable tool and preparing a compatible input because the corpora are often self-collected and unstructured. This paper describes ten popular open-source text visualisation tools for word-clouds. It is observed here that these tools take only a single text file as input while unstructured corpora have multiple-file format. So further in this paper a priority window technique is proposed to convert corpus into a small single text file that retains the characteristics similar to a standard model found in structured corpora. Experiments are performed over a self-collected corpus to demonstrate a real-life scenario from journalism application. The output word clouds show the effectiveness of proposed frequency-based technique. This technique is aimed at ease of use by users that lack expertise of data mining.

Keywords: text visualisation; summarisation; text mining; word-cloud; open-source visualisation tools; term weighting; feature weighting.

DOI: 10.1504/IJDATS.2023.132561

International Journal of Data Analysis Techniques and Strategies, 2023 Vol.15 No.1/2, pp.129 - 146

Received: 09 Jun 2022
Accepted: 11 Feb 2023
Published online: 28 Jul 2023 *