چكيده به لاتين
Abstract
Today, with the development of new technologies and the Internet and social networks, data production is constantly growing, and we are faced with the concept of big data. Therefore, we are facing many challenges in identifying useful information from this amount of data, storing, analyzing, and working with this data. On the other hand, data processing and analysis are not enough, and the human brain tends to find more efficient patterns through visual data representation. Considering an enormous challenge, one must also think about how to visualize this huge amount of data, which data to use, which visualizer to employ, what features to incorporate, and how to display it. The findings show that individual personality differences are highly consistent with users’ preference for information visualization and the use of adaptive visualization is effective in attracting users’ satisfaction. Also, the case studies have shown that the data-driven approach can lead to an interactive and user-centered visualization tool that strengthens the user’s understanding and insight and increases the power of visualization. The user-centered approach tries to optimize the design by placing the user at the center of the design and interacting and adapting to his needs. On the other hand, context-aware applications provide an opportunity to facilitate and strengthen human cognition with technological advances.
This research aims to design and prototype a data-driven User-centric data visualizer while paying attention to user context. We implemented this visualizer by presenting a comprehensive architecture. To achieve this, we investigated and evaluated the personality traits of 87 people (48 women and 39 men in the age range of 19 to 39 years), who were computer science students and graduates, using the NEO-FFI questionnaire. After examining the results, we selected three personality groups for the study. In the following, while interacting with users and observing some design principles and rules, prototyping three types of user interfaces according to the personality characteristics of users and one user interface for neutral mode was done. After comparing and evaluating several machine learning algorithms, we chose the logistic regression algorithm with a prediction accuracy of 0.87 for modeling. Also, the results of the final evaluation showed that users prefer the user interface designed according to their personality and needs. On the other hand, according to the context, a big step was taken toward increasing the level of user satisfaction in using data visualizers.
