Persuasive techniques are recently being explored by computer science researchers as an effective strategy towards creating applications that are aimed at positive attitudinal changes especially in the health domain but finding effective evaluation approaches for these technologies remain an herculean task for all stakeholders involved and in order to overcome this limitation, the Persuasive System Design (PSD) model was designed but researchers claim that the model is too theoretical in nature and some of its design principles are too subjective as they cannot be measured quantitatively. Hence, the focus of this paper is to critically review the PSD model and popular models currently being used to evaluate the usability of information systems as usability has been identified as an important requirement currently used to evaluate the overall success of persuasive technologies. To achieve the stated objectives, the systematic review method of research was done to objectively analyze the PSD model, its applicability as an evaluation tool was tested on a popular mobile health application installed on the Samsung Galaxy Tablet using android Operation system. Exhaustive evaluation of the application was performed by 5 software usability researchers using the method of cognitive walkthrough. From the analysis, it was realized that the PSD model is a great tool at designing persuasive technologies but as an evaluation tool, it is too theoretical in nature, its evaluation strategies are too subjective in nature and the 28 principles described in it overlap with one another. As a result, the PSD model was extended with an integrated usability model and the fuzzy Analytic Hierarchical Technique was proposed theoretically to evaluate usability constructs so as to make evaluation of persuasive technologies more quantitative in nature and easier for researchers to analyze their design early enough to minimize developmental efforts and other resources.

The data-driven methods capable of understanding, mimicking and aiding the information processing tasks of Machine Learning (ML) have been applied in an increasing range over the past years in diverse areas at a very high rate, and had achieved great success in predicting and stratifying given data instances of a problem domain. There has been generalization on the performance of the classifier to be the optimal based on the existing performance benchmarks such as accuracy, speed, time to learn, number of features, comprehensibility, robustness, scalability and interpretability. However, these benchmarks alone do not guarantee the successful adoption of an algorithm for prediction and stratification since there may be an incurring risk in its adoption. Therefore, this paper aims at developing a logical approach for using Empirical Risk Minimization (ERM) technique to determine the machine learning classifier with the minimum risk function for data stratification. The generalization on the performance of optimal algorithm was tested on BayesNet, Multilayered perceptron, Projective Adaptive Resonance Theory (PART) and Logistic Model Trees algorithms based on existing performance benchmarks such as correctly classified instances, time to build, kappa statistics, sensitivity and specificity to determine the algorithms with great performances. The study showed that PART and Logistic Model Trees algorithms perform well than others. Hence, a logical approach to apply Empirical Risk Minimization technique on PART and Logistic Model Trees algorithms is shown to give a detailed procedure of determining their empirical risk function to aid the decision of choosing an algorithm to be the best fit classifier for data stratification. This therefore serves as a benchmark for selecting an optimal algorithm for stratification and prediction alongside other benchmarks.

No abstract