Abstract:The management of customer services by telephone encounters several problems: an uncontrollable flow of calls, complicated resource management, a very high cost of service, and more. Opportunities to improve the quality of service, save time and money triggered the widespread implementation of artificial intelligence (AI) based callbot. This article outlines the straightforward workflow developed to model the architecture of the callbot. Therefore, several algorithms were evaluated and compared based on real knowledge of a call center of an insurance society. The algorithms considered are: k-nearest neighbours (KNN), support vector machine (SVM), random forests (RF), logistic regression (LR), and Na ̈ıve Bayes (NB). The comparison criteria are: correct responses, response time, accuracy, Cohen’s kappa and F1 score using n-gram (1.1) and (2.2). The results obtained show that the SVM (accuracy=70.29%) presents the best results on all the comparison criteria. The comparison between the results of the human agents and the callbot shows an improvement in several levels: the cost savings are greater than 80% on all the tests carried out, the holding time decrease to 0 seconds, and the processing time (almost a third or more). The results obtained sufficiently meet the objectives of this project.

Multi-agent system for customer relationship management with SVMs tool

Association Rules Mining Based on SVM and Its Application in Simulated Moving Bed PX Adsorption Process

Association rules extraction based on support vector machines

A Kind of Unusual Customers Recognition System Based on Multi-criteria Neural Network and CART in Telecom System

A process driven architecture of analytical CRM systems with implementation in bank industry

Design and Implement of Sale Management System Based on Multi-Agent in Pharmaceutical Enterprise

Application of Support Vector Machine in Bank Customer Classification

Application of data mining in travel agency customer relationship management system

Research on the CRM-oriented and Multi-Agent Based Data Mining Model

Sale and Allocation Management System Based on Multi-Agent Technology in Clothing Enterprise

Combination Classification Method for Customer Relationship Management

A Multi-SVR Based Fuzzy Modeling Method for Non-Stationary Time Series

Multi-sphere Classification Algorithm Based on 1-SVM

Design and implementation of bank CRM system based on decision tree algorithm

Decision trees combination method based on genetic algorithm with application in CRM

Customer Churn Analysis Based on Improved Support Vector Machine

Application of Multi-agent System in SCM

Customer Relationship Management Based on SPRINT Classification Algorithm under Data Mining Technology

Multi Agent System for Machine Learning Under Uncertainty in Cyber Physical Manufacturing System

Customers Behavior Modeling by Semi-Supervised Learning in Customer Relationship Management

Modeling of an artificial intelligence based enterprise callbot with natural language processing and machine learning algorithms