Qualitative usability feature selection with ranking: a novel approach for ranking the identified usability problematic attributes for academic websites using data-mining techniques

Abstract

Objective: The aim of this study is to identify common usability problematic patterns that belong to top-50 academic websites as a whole and then ranking of these identified usability problems is also provided. Methods:In this study, a novel approach is proposed that is based upon the integration of conventional usability testing and heuristic evaluation with data-mining knowledge discovery process. An experiment is conducted to evaluate ISO 9241-151 guidelines under 16-different categories by hundred participants who are frequent users of academic websites. After evaluation, the qualitative usability data is collected and different data-mining techniques i.e. association rule and decision tree are applied to recognize fully functional and problematic usability attributes. Identified problematic attributes represent common usability problems patterns related to academic websites from the qualitative viewpoint only. This study further prioritizes these problematic attributes by using the ranking algorithm that represents the order in which usability issues must be resolved. Results:In this study, 16-different categories are considered for usability evaluation of academic websites. The results show that no issues are identified in two-categories i.e. {Headings_Titles_Labels and The Home_Page}. In Scrolling and Paging category, horizontal scrolling is identified as a major issue whereas, in Internationalization category, the users do not identify supported languages on most of the academic websites. Users do not find websites to be highly secured under Security category. Our findings investigate that most of the issues are found in Search and Social Media categories. Furthermore, users easily locate 50.53% guidelines on websites as fully functional whereas, 49.46% of characteristics are considered as problematic usability features that are not functional on the academic website as a whole. Conclusions:Identification of common usability problems at an early stage can lower substantially the development efforts in cost and time. Software developers can restrain from these potential usability problems during the development of novel systems under the same context. Providing appropriate solutions for these problems can become valuable in software development. The proposed approach concludes Sagar and Saha Hum. Cent. Comput. Inf. Sci. (2017) Page 2 of 24 Sagar and Saha Hum. Cent. Comput. Inf. Sci. (2017) 7:29 that conventional usability evaluation methods can go beyond just than testing of systems. The study is a milestone towards identification and prioritizing problematic usability features for academic websites and helps in providing the wholistic approach of usability problematic patterns for web-domain. RESEARCH

QER: a new feature selection method for sentiment analysis

“…Feature selection methods are used to rank features so that non-informative features can be removed to improve the classification performance [18]. Some researchers have investigated the effects of feature selection for sentiment analysis [3,[8][9][10][19][20][21][22][23][24][25]. For example, Yang and Yu [3] examine IG for feature selection and evaluate its performance using NB, SVM, and C4.5 (popular implementation for DT) classifiers.…”

Manual versus automated qualitative usability assessment of interactive systems

“…12 Besides, ISO 9241-151 usability guidelines *Top Fifty Websites mentioned in National Institutional Ranking Framework are assessed under sixteen-different categories are also evaluated but by involving the end-users. 23 Association rule and decision tree techniques are administered to find functional and problematic usability features. Further, these features are prioritized by employing another ranking algorithm.…”

“…Qualitative usability assessment 8 is also done to evaluate any interactive system for a particular context of use. Further, some articles are identified that present novel study for usability assessment with data‐mining techniques like association rules and decision rules 7–9,12,23 …”

Feature selection algorithm for usability engineering: a nature inspired approach

“…Over the years, feature selection has been used to reduce features in diverse fields, for instance,in health care data analysis [32], flash flood hazard assessment [33], and for medical image diagnosis [34], among other applications. It has also been used in the area of software usability, for reduction of usability attributes [8,9,11], to identify problematic usability attributes [35] and to detect usability deficiencies by monitoring the amount of time taken in different tasks by users [36].…”