Development of a Comparative Table Creation Support System
Takayuki Arita (AY 2013)
Nowadays, searching for information on the Web has become an everyday activity. There are two types of information retrieval behaviors: reference-type information retrieval, which is performed when the purpose is clear, and exploratory information retrieval, which is performed when the purpose is not clear, such as when searching for a topic for a thesis. One of the purposes of exploratory information retrieval is to "compare" multiple objects, and comparison tables are particularly common on the Web. A typical comparison table is organized around attributes that are common to the objects being compared. Comparison tables are also thought to influence the task of making selection decisions. While comparison tables are useful, they require users to create their own comparison tables for topics that do not exist on the Web, which is time-consuming. There are several systems that assist users in creating comparison tables, but they all have the drawback that it is difficult for users to enter comparison targets and attributes when their information needs are ambiguous. The purpose of this study is to construct a system that supports the creation of web-based comparison tables and to investigate whether comparison tables have an impact on the aforementioned decision-making tasks.
n this study, we used the UCD (User Centered Design) method, which is a method of designing from the user's point of view, to construct the system. The system was developed as a web application. Next, the system presents candidate words for attributes, and the user selects the attributes to be added to the table, and then fills in the table contents. Then the system presents candidate attributes and the user is asked to select the attributes to be added to the table and fill in the table contents. The system creates the table by repeating these steps. In the human subject experiment conducted to evaluate the system, the topic of the comparison table was changed in Condition I, in which the subjects were asked to use Google Search and Excel, and Condition II, in which the subjects were asked to mainly use the system constructed in this study, and in each condition the subjects actually created the comparison table once. After the work, the subjects were asked to answer a questionnaire.
Based on the results of the questionnaire, it was found that condition II was superior to condition I in "the stage of selecting the comparison target in the stage of creating the comparison table" and "the stage of entering the table contents". It was also found that the number of attributes in the comparison tables created under Condition II was greater than those created under Condition I.
Although it was not possible to prove whether the comparison table affected the decision task, we proposed eight suggestions for improving the system based on the results of the experiment.
As a future direction, we will implement the proposed system improvements and conduct the subject experiment again, following the UCD process. Then, we will evaluate the system and investigate its impact on the decision task.
(Translated by DeepL)