Title: Mining customer reviews to evaluate the contact centre agent performance using custom kernel functions

Authors: A. Santhosh Kumar; Punniyamoorthy Murugesan; Ernest Johnson

Addresses: National Institute of Technology, Tiruchirappalli – 620 015, Tamil Nadu, India ' National Institute of Technology, Tiruchirappalli – 620 015, Tamil Nadu, India ' University of Regina, Regina, S4S 0A2, Saskatchewan, Canada

Abstract: In today's digital world, the exponential growth of unstructured text data necessitates businesses to rethink their organisational strategies based on the insights extracted from data using text or opinion mining. To extract opinions from text documents, various machine learning algorithms are utilised, with support vector machine (SVM) being a popular one due to its ability to efficiently classify nonlinear data using the Kernel trick (Kernel function). This function implicitly transforms the input to a higher dimensional vector space, making it easier to classify data linearly. In our study, we have applied the dissimilarity kernel function, which is suitable for sparse data. We evaluated the performance of the new kernel function in classifying opinions from customer feedback in the business to consumer (B2C) contact centre industry and ranked contact centre agents based on the customer feedback data.

Keywords: opinion mining; Jaccard dissimilarity kernel; custom kernel functions; contact centre agent performance.

DOI: 10.1504/IJENM.2024.140524

International Journal of Enterprise Network Management, 2024 Vol.15 No.3, pp.245 - 260

Accepted: 01 May 2023
Published online: 22 Aug 2024 *

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