Development of identity attribute metrics model based on distance metrics.
Loading...
Date
2020
Authors
Kabwe, Felix Musama Lameck
Journal Title
Journal ISSN
Volume Title
Publisher
The University of Zambia
Abstract
The growth in the use of online services on the World Wide Web has proliferated into cyber mischief, personality or object misrepresentation, and cybercrime. Diverse entities of different interests and intentions form a wide range of complex online identities and beneficiaries of the online activities. Fraudsters and criminals hide their online identities to steal services, assets and other valuables or harm innocent internet users. This research would help in strengthening of identity management systems as a way to arrest this growing problem and guarantee secure online services and online interactions. This research desires to identify a mathematical model that would help in improving cyber security in digital identity management. This work intends to develop metrics models based on distance metrics in order to quantify the credential identity attributes used in online services and activities. This study adds knowledge to past work on the subject matter to provide quantitative analysis to quantify the credential identity attributes in online services. The study considers major sources of identity attributes currently being used in the application and registration forms for the various services offered both in cyber and real space. The study further explores the extraction of key identity attributes that were extracted from identity tokens like identity documents, application and registration forms for the various services offered both in the cyber and real space. At the core of the research, the study seeks to establish how we would develop the identity attribute metrics model which could be used to quantify the identity attributes based on distance metrics mathematical models. The study utilized survey research with closed-ended researcher administered questionnaire. A total of 160 questionnaires were administered with a response rate of 93%. The primary data obtained from questionnaires was analysed using Statistical Package for Social Science (SPSS) and Excel. The respondents were drawn from Banks (14%), Churches (12%), Government of the Republic of Zambia (6%), Hospitals (16%), Insurance (10.7%), Mobile Phone companies (2%), and less than 1% from Pensions. Others were Schools (21%), Universities (16%), and Utility companies (1.3%). The techniques that have been used include data mining techniques and statistical analysis. The perception constructs in the research included Usefulness, Trust, Ease of use, Image, and User satisfaction. It was observed that some attributes were more important than the others in identifying entities. Statistical analysis revealed that among the constructs that were used, Usefulness, Trust and Ease of use were strongly related. Tools to text mine the identity attributes helped to generate statistical data to come up with a quantitative model metrics to assist in the identification of an online entity. Using a detailed literature review, questionnaire surveys in this area, text mining of the identity attribute from the
application forms, and the results of the study helped to develop the identity attribute metrics model. An identity attribute metrics model based on distance similarity has been proposed. The Distance similarity is based on Cosine Similarity measure. Based on this study, digital identity management in online services and activities should adopt the developed Cosine Similarity measure, the identity attribute metrics model based on distance metrics, to strengthen online identity management. This will help to curb online fraud, identity theft, and other cybercrimes. This model could be augmented to past efforts to come up with a multimodal solution and add value to the resolution of the said problem.
Description
Thesis of Master of Science in Computer Science