Eugene Santos, Jr.*, Hien Nguyen+, Fei Yu*, Deqing Li*, John T. Wilkinson*,
Thayer School of Engineering
Hanover, NH 03755
+University of Wisconsin-Whitewater
Dept. of Math and Computer Science
800 W. Main Street
Whitewater, WI 53190
With the increasing availability of online resources, collecting information on the Web and analyzing data play important roles in today’s problem solving task. 1.Baum L. E., Petrie T. , Soules G. , and Weiss N. A maximization technique occurring in the statistical analysis of probabilistic functions of Markov chains. The Annals of Mathematical Statistics, 41 (1), pp. 164-171. 1970. It has be found that a user’s cognitive styles affect their searching/browsing behaviors, their assessment of the relevancy of a web page, and their decision making processes. However, little research has been conducted that explore the impacts of a user’s cognitive styles on an analytical process. The challenges here are three fold: First, popular information retrieval and filtering systems on the Web do not take into account a user’s cognitive styles even though this factor is known to affect a user’s information seeking behaviors, and a user’s assessment of text summarization. Secondly, the unavailability of well-defined and relevant testbeds poses critical challenges to the process of capturing the goal-oriented and compound nature of an analytic process. Third, the open and diverse nature of the Web creates uncontrollable noise/influences such as environmental factors (variations in interfaces, dynamics of web information, etc.) or even credibility of participants that can affect the results of such a study. Therefore, in this paper, we explore the problem of the impacts of a user’s cognitive styles on analytical processes in the domain of intelligence analysis. Our results can also be used and extended to the Web community to solve analytical problems done in Web settings. 2.Blei, D. M.; Ng, Andrew Y.; Jordan, Michael I; Lafferty, J. Latent Dirichlet allocation. Journal of Machine Learning Research, 3,pp. 993–1022. 2003 Given a particular analysis task, it is likely that different people will take different approaches to solving the task. For example, political analysts will vary in how they perform their assessments of tasks such as “What will be the political repercussions of passing health care reform legislation?” Some will delve very deeply into the details of how individual politicians will be affected in upcoming elections by specific elements of the legislation such as abortion; while others will examine the national picture of how the political landscape is expected to change in the long run. As such, we are interested in studying the impact of a user’s cognitive style on intelligence analysis tasks. Understanding a user’s cognitive style is critical to determining their preferred methods for reading, remembering, learning, perceiving, and searching for information. Cognitive styles have been found to affect a user’s information seeking tasks, the ways users interact with a graphical user interface, and a user’s reading task. In our previous work, we discovered that there exists a consistent correlation between an analyst’s actions and their final analytical conclusions from task to task. 3.Chen, Y. S. and Macredie, D. R. Cognitive styles and hypermedia navigation: Development of a learning model. Journal of the American Society for Information Science and Technology, 53(1), pp. 3-15. 2002. In this paper, we advance this effort further by investigating how a user’s cognitive style may affect his/her analytical process. This involves two challenging steps: First, we need to determine a user’s cognitive style in an analytical process automatically by analyzing the sequence of actions a user takes. Second, we need to...