Implementing a Benchmarking Model for Assessing Internet Users’ Complaints in University of Ibadan, Nigeria

Abstract

The Internet is the most important research tool in a university, research being a major activity. It is therefore important to capture users’ satisfaction of the Internet services of a university. This work seeks to improve the performance of the Internet as a research tool by developing a benchmar king model of users’ complaints using University of Ibadan as a case study. Python Natural Language Process (NLP) technique was used to analyse users’ complaints while tokenization, stemming and stop words were used to extract keywords. The identified pro blems from the complaints benchmarking were classified into 32 groups which are clustered into four separate unit which are: “No signals”, “weak signals”, “physical problems” and “users related’ problems. The user related problems had the highest count, fu rther broken into, “Software problems”, “Hardware problems” and “Network problems”. The study revealed that users’ problems have the highest constraint within the university network . A database was created to capture and store user network complaint which was analysed using natural language processing algorithm . Similar complaints were matched with solutions, in order to improve on users’ satisfaction while using the Internet as a research tool as well as to facilitate the work of network administrators in decision making.