Machine Learning-Based Sentiment Analysis of Incoming Calls on Helpdesk
Open Access
- 27 October 2023
- journal article
- research article
- Published by Auricle Technologies, Pvt., Ltd. in International Journal on Recent and Innovation Trends in Computing and Communication
- Vol. 11 (9) , 21-27
Abstract
In today's daily life we are getting so many anonymous calls. Some calls are related to loan marketing and finance. As per the survey, one person is getting 26% spam calls in a day. The proposed methodology accepts user calls and based on the conversation the spam numbers are identified and the same information is provided to the other callers. This is possible because of machine learning-based sentiment analysis. Sentiment analysis is the subdomain of machine learning. The goal of this research is to propose an adaptive methodology for incoming calls. The sentiment-based incoming calls help desk works with freely available lexical resources WordNet, SemCor, and OMSTI. The discussed methodology accepts user conversations in audio format the speech-to-text conversion of the audio will be done. After pre-processing the keyword is detected from the statement. The word2Vec word embedding technique is used for representing words from document space to vector space. The 150-200 dimensional word vector is generated. The WordNet is used for sense mapping and keyword identification. Based on the sentiment analysis of input calls the decision is taken whether to accept or reject calls. This methodology is generating superior results for supervised machine learning models.Keywords
This publication has 6 references indexed in Scilit:
- Detecting Sentiment Polarities with Comparative Analysis of Machine Learning and Deep Learning Algorithms2023
- Sentiment Analysis Perspective using Supervised Machine Learning Method2023
- A novel multi-scale CNN and attention mechanism method with multi-sensor signal for remaining useful life predictionComputers & Industrial Engineering, 2022
- Word Sense Disambiguation for Large Documents Using Neural Network Model2021
- LSTM-based Deep Learning Model for Stock Prediction and Predictive Optimization ModelEURO Journal on Decision Processes, 2021
- Meta-analytic support vector machine for integrating multiple omics dataBioData Mining, 2017