A Study On The Measurement Of Users’Mental Model For Website Product Catalog

Our study showed that, compared to the other two types of sites:search engine sites, digital libraries, categories of e-commerce site take more attention. However, e-commerce website product catalog exists the difficulties of users understanding the web interface, the problem of low efficiency when users using the website. The problem dues to the difference between web interface performance model that based on designer’s mental model and the users’mental model based on the understanding of interface. In the transmission of website information, reducing the differences between the user’s mental model and website’s performance model is the key to improve efficiency of the users’search performance, and the premise of reducing the differences is the measurement of the differences. Therefore, how to measure the difference between the web site performance model and the users’ mental model for e-commerce website product catalog is the fundamental problem of this study. The following is the main research work of this paper.(1) The measurement method of mental model and think of its application design. Combined with the task of our study, we thought of the application of the methods. The methods include concept-mapping method, the conceptual similarity calculation methods and Bayesian network method. It provides the basis for making a reasonable choice of measurement method for the experiments.In which, the concept of similarity calculation is an important element of this study to explore. The basic idea is using the method of concept map to obtain the users’mental model, and then calculate the similarity with performance model of website, thereby measuring the differences between the user’s mental model and performance model of website. For that, we explored the clustering method, path search method, and document similarity calculation method, semantic similarity calculation method based on the concept of tree and string similarity calculation method. After a series of analysis and comparison, clustering method and the string similarity calculation method eventually served as the measurement method to measure the difference between the mental model of users and the performance model of website. One of which is the main innovations of this thesis.(2) Theoretical analysis of the website product catalog search mechanism. We explored the classification search mechanism, analyzed the cognitive processes and the action mechanism of their mental model of the classification specifically. And we analyze the principle of the difference between performance model of website and the mental model of users. This provides a theoretical basis for the experimental design, measuring experiments of mental model behind also validate the mechanism, and this is one of the innovations of this thesis.Thus we proposed theoretical assumptions.1) Hypothesis1:When users browse categories node, they were comparing the node concept with the mental model concept, the degree of the difference between initial model of users’and performance model of website is greater, and the efficiency of searching is lower.2) Hypothesis2:Based on the search target, when users search on the website, the user will revise their mental model based on classified nodes, to fumble near the path of standard classification that actual stored the search target.3) Hypothesis3:Classification node selection is depending on the results that node concept and the concept of mental models were matched.4) Hypothesis4:Based on the search target, search results determine whether the user will revise mental model. If they can’t find the target, users will revise their mental model, make it get closer to the standard classification path; if they find the target, users will maintain their mental model, and make the experience to later search activities. Searching product from the beginning to the end, the differences between model of users’and performance model of web site were getting smaller and smaller.(3) The user’s mental model measurement experiment for website product catalog. Firstly, we used concept-mapping method to record the mental model that participants understanding of goods, and to explore extraction and quantitative characterization of the performance model of website. And we used clustering method to cluster mental models of users and performance model of website that mixed together, to analyze the differences between them, and comparative analysis of test search performance, the final experimental results show that the establishment of the above theoretical assumptions, we verified the theoretical Hypothesis1。(4) The experiment about users’ mental model dynamic measurement for website, this is the main innovations of this thesis.We designed experimental platform based on the website product catalog search mechanism and dynamic Bayesian network. We gave a few pieces branch classification trees to search for goods that they are related to the e-commerce website. We asked subjects to search for the product "electronic dictionary", and this product was attributable to one of the categories branch. We asked subjects to write about a all changing concept map of mental model in the search process. We asked subjects to browse classification nodes and give classification nodes’ expectation. At last, the experimental results verify the proposed four hypotheses of the theoretical part.