ABSTRACT
Numerous websites in this contemporary time have been plagued with many usability issues which have hitherto made the websites not effective and efficient for users while searching for information. Consequently, different website usability evaluation models have been proposed to help in evaluating websites. However, most existing models are rather too ambiguous and not easy to use. Also, selecting and ranking websites based on usability with respect to numerous criteria have become a very important decision-making process among users. Additionally, there is no existing machine leaning model developed to classify websites usability based on user’s rating due to lack ofusability ratings data. This thesis therefore proposes a new integrated usability evaluation model using Fuzzy Analytical Hierarchy Process (FAHP) with Artificial Neural Network (ANN). Five criteria of Speed (S%a). Navigation (N), Ease-of-use (Eo), Content (C%a) and Aesthetic (A) obtained through factor extraction out of initial seven criteria proposed are used in the study. Six Nigerian universities websites with good webometrics ranking are used as alternatives. These are University of Ibadan (UI), Covenant University (CU), Obafemi Awolowo University (OAU), University of Nigeria Nsukka (UNN), University of Lagos (UNILAG) and Ahmadu Bello University (ABU) websites. Two sets of usability data were collected via google forms from 233 and 169 participants. Results from FAHP indicates that UI website has the highest global priority weight and hence is ranked as number one. This is followed by CU, OAU, UNILAG, UNN and ABU websites respectively. Also, final criteria weights obtained are 0.321S%, 0.208N~, 0.197E%6, 0.166C% and 0.108Aa, respectively. This implies that the first and most important criteria to website users is speed. Weights obtained from FAHP model were preprocessed and used to train six machine learning algorithms which are Artificial Neural network (ANN), Random Forest (RF), Decision Tree (J48), Simple Logistic regression (SLOG), Bayesian Network (BaNET) and Logistic Model Tree (LMT). Results show that ANN has the best overall performance with accuracy (Ac) of 93.36% while RF, LMT, SLOG, J48 and BaNET have 90.12% A, 88.09% A&, 88.18%A, 88.18% Ac and 83.63% Ace respectively. The FAHP model is further integrated with ANN to classify the user’s websites usability ratings. The ANN structure is 5-3-1 with logsig and trainbr as activation and transfer functions respectively. The best performance was obtained at learning rate (l) of 0.8, momentum (m) of 0.9 and threshold value(h) of 0.59. Further results obtained shows a precision (P), recall (Rae) and F-measure (Fame) values of 98.44%P, and 95.45%Rand 0.96F~ respectively. It is recommended that this integrated model, which can be used for users’ websites usability evaluation, ranking and prediction be adopted by IT practitioners and web developers.
CHAPTER ONE
1.0 INTRODUCTION
1.1 Background to the Study
The internet is nowadays a major source of information through the use of websites, and as a consequence, websites generally have been serving as information gateway to different types of organisations (Dingli and Cassar, 2014; Esmeria and Seva, 2017; Monzer, 2015; Sun et al., 2017). Nowadays, websites offer an easy means of searching and retrieving information about any kind of organisation. Basically in this information technology era, the first impression users of websites have about organisation is virtually based on the look of its website (Ismailova and Kimsanova, 2017). According to the Internet World Statistics (Internet World Stats, 2020) population of internet users in the world is now over 4.5 billion from 360 million in 2000 with 58.81 % penetration rate. Following a similar rate, in Nigerian internet users’ population have grown from two hundred thousand (200,000) in the year 2000 to over one hundred and twenty-three million (123,000,000) as at June, 2020 with 61.4. 7% penetration rate. All these point to the fact that accessing different types of website is inevitable and a must task for these billions of different users in today’s information technology driven world.
Consequently, websites have become an essential tool for many organisation because of its wide reach, broad acceptance and general capability to share information. Till date millions of websites have been created and developed and there exist every kind of websites varying from easy to difficult-to-use (Dominic et al., 2013; Rajapaksha and Fernando, 2016). In addition to this, vital roles are being played by the web in the diverse domains of business, education, industry, agriculture, health and entertainment among others. Hence, the degree of website usability and quality coupled with its development has been a major concern to usability researchers (Almahamid et al., 2016; Djordj et al., 2013; Manzoor et al., 2012; Mvungi and Tossy, 2015) Different genre of websites exists and each is suitable for a particular audience or purpose. Among these are academic websites for educational institutions like universities, polytechnics, colleges and specialized institutions. Other genre includes e-commerce websites, hotel and tourism websites, airline websites, e-government websites, banking websites, political party websites and many others. For academic institutions, their websites are meant to provide information to a wide range ofusers which include prospective and enrolled students, staff, parents, institutional ranking bodies as well as other categories ofusers. These websites not only serves as a platform for the stakeholders to exchange information, they also serve as communication tools and help to shape its image (Mentes and Turan, 2012; Abdallah and Jaleel, 2015; Galovicova et al., 2016).
Today, millions of people are searching for information on university websites annually. These includes, prospective students looking for schools on potential courses available, subject experts, fees information among other vital information (Affandy et al., 2017; Alahmadi and Drew, 2016; Jati et al., 2018) Enrolled students search for course information, lecture location, materials and times, account access, results updates, schools’ calendars, fees payment, news update, teacher’s information. Prospective applicants may search for job prospect, vacancies, available facilities, research output, funded projects, sample thesis and project. The main underlining issue is that users should find what they are searching for easily and the content should be easy to understand (Sarsarabi and Sarsarabi, 2015).
In Nigeria, there is increasing competition among the universities especially with respect to web visibility ranking. At present, there are a total of 172 (one hundred and seventy two) universities comprising forty four (44) Federal universities, forty nine two (49) States universities and seventy nine (79) private universities respectively as released by National University Commission (NU C, 2021) In the latest webometric ranking of higher institutions, there is no university in Nigeria among the top one thousand (1000), while twenty six universities are in the category of top five thousand (5000), fifty six (56) appear in the top ten thousand (10, 000), while the rest are in the rank of between ten thousand (10000) and twenty three thousand (23, 000) out of the total of twenty three thousand, three hundred and sixty eight (23,368) institutions worldwide that were included (Cybermetrics Labs, 2020). Nevertheless, while this statistic is not encouraging, there is still greater web presence among Nigerian universities than what was obtained in the past. It is therefore necessary to see how this can improve over time.
Having a good web presence will make potential users to know about the school and this will in tum attract many more visitors to the school websites. As a result, all universities therefore will strive as much as possible to have a user-friendly website which are both functional and usable. Due to increased competition, universities seek to attract the best of all students, faculty and research grants. Hence, there is dire need to increase the web visibility of each university websites (Kargar, 2012; Okello-Obura, 2015; Peker et al., 2016). To achieve this, there have been several attempts in rebranding and redesigning of websites by various universities administration. All these are with the aim of making their websites accessible, usable and have positive impact on users. Hence, the need to improve on the usability and quality of these websites so as to prevent users from being frustrated in this information age. Also, in this competitive era, if users cannot find what they are looking for, they will simply tum to competitor’s websites. This necessitates the need for a good usability (Manzoor et al., 2019).
In this information age, users of any websites are mostly concerned with two major issues finding the information being sought with ease and finding it in a timely fashion. To achieve this, a high level of usability which is one of the important criteria in measuring website quality is required (Aziz and Adzhar, 2015; Roy et al., 2016). According to International Standard Organisation, ISO 9241-11, usability can be defined as “the extent to which a product, service or system can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context ofuse”. It is further defined as the effectiveness, efficiency, and satisfaction with which specified users achieve specified goals in particular environments (Speicher, 2015).
From websites context, usability is seen as an important attribute of quality which describes how easy it is for users to navigate through the website. It can be viewed as the extent to which a goal can be achieved by users successfully by learning and using websites. As earlier stated, a functioning website is needed by every organisation for easy information dissemination to the public. In this context, university websites as specialized genre ofwebsites are supposed to be given adequate attention in terms of usability due to numerous services to its users worldwide (Yerlikaya and Durdu, 2017). However, many existing websites have been discovered over the years to have usability problems (Arasid et al., 2018; Stoimenova and Christozov, 2013). This has consequently led to the growing interest by researchers to develop users’ models to measure and evaluate website usability so as to fully discover its inherent problems. In further response to this, there have been increasing attention in website usability evaluation model research in the field of Human Computer Interaction (HCI) (Leung et al., 2016; Nagpal et al., 2016b; Presley and Fellows, 2013) . In HCI, usability of interfaces is being considered a factor of growing importance in application development, especially in web-based application. According to Peker et al. (2016), usability of websites is one of the popular subjects in the HCI literature which focuses on the interaction between people and Information & Communication Technologies (ICTs). Several research efforts have shown that usability is one of the most important issues in ICTs (Affandy et al., 2017; Das and Patil, 2014; Mvungi and Tossy, 2015; Nagpal et al., 2016a; Niazi et al., 2020). Till date, one of the challenges faced by HCI researchers is how best to measure or evaluate website usability.
As a result of this, several researchers have proposed different models for website usability evaluation. Most of these models are based on inspection methods and formal experimental test which are generally known as the traditional approach (Affandy et al., 2017; Hussain and Kadhim, 2014; Ismailova and Kimsanova, 2017; Majrashi and Hamilton, 2015; Nagpal et al., 2016b; Subair and Aleisa, 2016). However, in usability there are several criteria involved and determining which one contributes more to usability and at the same time ranking the alternative websites based on the criteria is a complex decision-making process. This therefore requires the formulation of websites usability problem by using a Multi-Criteria Decision Making (MCDM) approach. Website evaluation hence, belongs to MCDM field which involves making a preference decision, such as evaluation or selection over the available alternatives using a set of criteria. In MCDM several alternatives are usually involved, among which the decision-makers (DMs) have to give weights to each criterion (Jain et al., 2016; Ozkan et al., 2020) Also, with the advent of machine learning, attempt to use machine learning techniques in usability evaluation research have achieved little or no success (Boza et al., 2014; Korvald et al., 2014; Nayebi, 2015; Oztekin et al., 2013; Sagar and Saha, 2017) This approach involves using different machine learning algorithms like Neural network, support vectors machine, decision tree, linear regression and the likes to generate and model users usability data. This can then be used for prediction and consequently give a better insight into usability data. This has however suffered several limitations partly because of the nature of data that is required for machine learning training and the low performance output of the machine learning algorithms used in the models (Korvald et al., 2014; Sagar and Saha, 2017; Taj et al., 2019).
Therefore, to handle the dual problems highlighted above with a view to getting better insight into usability data from users’ perspective and further help in usability users rating prediction, the need arises to integrate machine learning techniques with MCDM approach. This combined data-based and expert-user based approach is the main focus and contribution of this thesis. This research therefore is based on integrating an MCDM approach based on fuzzy Analytical Hierarchical Processing (AHP) with Artificial Neural Network (ANN). This integrated approach handle both the subjective and objective aspect of usability evaluation thereby eliminating biases exhibited by human being during evaluation. More so, appropriate ranking of websites performance based on usability as well as better user website usability rating is also achieved.
1.2 Statement of the Problem
Usability is a key factor in the quality and success of a website. This is because the ease, comfort, distraction or difficulty that users experienced with websites determines their success or failure (Hasan, 2013; Hasan and Morris, 2017; Quinones and Rusu, 2017). Most times, a number of users experience frustration due to the fact that the information been sought for on the websites are not readily available or requires great efforts to access simply due to usability and accessibility problems in websites (Jano et al., 2015; Manzoor et al., 2019; Sagar and Saha, 2017). At present there are many usability issues with most academic websites and the major challenge is to know the appropriate usability issues to tackle in order to ensure better usability. If a website does not meet user expectations with an appropriate level of usability, it will lead to increase in website failure rate. As a result, users’ ratings about the website will be poor (Esmeria and Seva, 2017; Nagpal et al., 2016a; Yerlikaya and Durdu, 2017). Though attempts have been made by researchers to identify different criteria of website usability in the academic field, there is yet to be a widely acceptable model (Kaur et al., 2016a; Quinones and Rusu, 2017; Subair and Aleisa, 2016)
More so, most studies focused more on website quality criteria but only a few focused-on website usability especially in academic domain and such are not adequate considering its relative importance. Also, providing a machine learning model for website usability evaluation especially for academic websites is also a great challenge to researchers in the field of HCI. Most of the existing usability evaluation models have been using the traditional approaches and do not really solve the usability issues (Dingli and Cassar, 2014; Hasan, 2013; Jano et al., 2015; Subair, 2014). Furthermore, classifying website usability based on users rating is non- existence and this is very important with the advent of data mining. This implies that, there is no existing model that can aid in classifying and predicting user rating based on website usability so as to know the class a particular website belongs based on some criteria or parameters. So, the need arises to develop better models which are clear, simple, easy to use and can in in users’ website usability prediction with good performance.
1.3 Aim and Objectives of the Study
The aim ofthe study is to develop an Integrated Fuzzy AHP and ANN Model for Website Usability
Evaluation. The aim will be achieved through the following objectives.
I. To identify and formulate a hierarchy ofcriteria for academic websites usability evaluation.
II. To develop a fuzzy AHP model based on the criteria identified above to determine and measure the weight of the usability criteria.
III. To carry out comparison evaluation on the data obtained from model (ii) above using different machine learning algorithms.
IV. To integrate the model with Artificial Neural Network for users’ website usability rating classification.
V. To evaluate the performance of the integrated model using standard machine learning performance evaluation metrics.
1.4 Significance of the Study
The website of a university gives the first impression about the school, it is therefore very essential for each university to create a usable, visually attractive and appropriate web presence (Ismailova and Kimsanova, 2017). Poor usability often means poor user interaction and hence reduced user acceptance and satisfaction. Due to neglect of usability issues, a lot of time, efforts and money are being wasted from time to time on redesigning academic websites in many educational institutions. The intention to continue or quit browsing a website depends on the first impression with the website A website that is acceptable will be judged by users within a minute and if they are not satisfied with the content, the websites will be discarded (Ulutas, 2019). This may force some potential students and faculty to abandon the websites if the required information is not readily available. This study will be of immense benefits to users of academic websites, the management as well as web designers of various academic institutions.
Users will find it very easy to retrieve required information effectively, efficiently and satisfactorily while reducing cognitive load. The management of the institution will also benefit by spending less money, efforts and time on rebranding and redesigning websites on regular basis. Web designers on the other hand will be able to know which area(s) of the websites need improvement and attention so as to improve the usability. Therefore, knowing the important criteria that influence usability is very important as it will help the stake holders to pay attention to factor(s) with the highest weight and then identify the best way to improve it (Roy et al., 2014).
1.5 Scope and Limitation of the study
This research covers only the usability aspect of websites which is a very important component of website quality. The data collected covers users’ interaction with six identified university websites with good webometric presence. Users testing used are both the moderated and unmoderated which include also laboratory test conducted during the different phases of the study. The class of the ANN is a binary class based on user’s evaluation. The target audience of the study are enrolled students; both undergraduates and postgraduates.
This study only takes into consideration the usability aspects of human computer interface; in this case university websites. Though the study can be extended to other genre of websites but the target users are mostly users of academic websites. Also, the last phase study is limited to laboratory setting where the users’ activities with the websites in use can be easily observed for authentic and adequate data collection. The choice of machine learning algorithms used in performance evaluation is limited to those with high accuracy.
This material content is developed to serve as a GUIDE for students to conduct academic research
INTEGRATED WEBSITE USABILITY EVALUATION MODEL USING FUZZY ANALYTICAL HIERARCHY PROCESS AND ARTIFICIAL NEURAL NETWORK
A1Project Hub Support Team Are Always (24/7) Online To Help You With Your Project
Chat Us on WhatsApp » 09063590000
DO YOU NEED CLARIFICATION? CALL OUR HELP DESK:
09063590000 (Country Code: +234)
YOU CAN REACH OUR SUPPORT TEAM VIA MAIL: [email protected]
09063590000 (Country Code: +234)
