Zegu Adverse Behaviour And Protest Prediction System

ABSTRACT

This document entails full details of research given in phases of the development of the Adverse behaviour and protest prediction system for Zimbabwe Ezekiel Guti University (ZEGU). It is a sentiment analysis system designed to read sentiments in text posted by students and members of the university community at large in order to aid the institution’s administration in predicting possibility of protests at the university. It was seen necessary to develop this system due to several problems resulting from unforeseen protests at the institution as there was no way of reliably assessing the general mood in the student community. After the system was idealised and put forward for development, a feasibility study was conducted in various aspects namely economic, technical, social and operational feasibility in order to assess the viability of the system. The study on feasibility of the system proved that it was rational to embark on the development project and thus current system information was gathered using various methodologies namely interviews, questionnaires and records inspection. Analysis of the existing system commenced taking into account weaknesses that needed to be addressed. Further emphasis was put on the requirements of the new system in terms of functional and non-functional. Risks that were present to the project were outlined and noted with possible solutions given for them. Design of the system was then done with several Unified Modelling Language diagrams employed to represent the system and its data in different dimensions. A number of design aspects were discussed within the document which included physical, database, program and interface design. Visual illustrations of the designed concepts discussed were given across all design aspects to enhance understanding of the system. Security design was then explained in different dimensions and interest areas such as physical, operational and network security in relation to the system. Implementation of the developed system then came under discussion, clearly outlining the procedures and strategies used in testing and implementing the system. Various testing techniques were applied to the system and subsequently the changeover strategies considered for delivering the system were explained with the recommended strategy being noted. The system maintenance plans available for the system together with the conclusion given on recommending the one to adopt was given and this was the adaptive maintenance to make sure that the system evolves along with the everchanging technology in the modern age. The system recommendations for improvements in future were listed such as integrating it to other social sites for data analysis in order to broaden the opinion base for the institution.