Abstract:
In this article, we report our implementation and comparison of two text clustering techniques. One is based on Ward's clustering and the other on Kohonen's Self-organizing Maps. We have evaluated how closely clusters produced by a computer resemble those created by human experts. We have also measured the time that it takes for an expert to `clean up' the automatically produced clusters. The technique based on Ward's clustering was found to be more precise. Both techniques have worked equally well in detecting associations between text documents. We used text messages obtained from group brainstorming meetings.
Abstract:
This article describes research in the application of a Kohonen Self-Organizing Map (SOM) to the problem of classification of electronic brainstorming output and an evaluation of the results. Electronic brainstorming is one of the most productive tools in the Electronic Meeting System called GroupSystems. A major step in group problem solving involves the classification of electronic brainstorming output into a manageable list of concepts, topics, or issues that can be further evaluated by the group. This step is problematic due to information overload and the cognitive demand of processing a large quantity of textual data. This research builds upon previous work in automating the meeting classification process using a Hopfield neural network. Evaluation of the Kohonen output comparing it with Hopfield and human expert output using the same set of data found that the Kohonen SOM performed as well as a human expert in representing term association in the meeting output and outperformed the Hopfield neural network algorithm. In addition, recall of consensus meeting concepts and topics using the Kohonen algorithm was equivalent to that of the human expert. However, precision of the Kohonen results was poor. The graphical representation of textual data produced by the Kohonen SOM suggests many opportunities for improving information organization of textual information. Increasing uses of electronic mail, computer-based bulletin board systems, and world-wide web services present unique challenges and opportunities for a system-aided classification approach. This research has shown that the Kohonen SOM may be used to automatically create "a picture that can represent a thousand (or more) words.".
Abstract:
As a result of growing misuse of online anonymity, researchers have begun to create visualization tools to facilitate greater user accountability in online communities. In this study we created an authorship visualization called Writeprints that can help identify individuals based on their writing style. The visualization creates unique writing style patterns that can be automatically identified in a manner similar to fingerprint biometric systems. Writeprints is a principal component analysis based technique that uses a dynamic feature-based sliding window algorithm, making it well suited at visualizing authorship across larger groups of messages. We evaluated the effectiveness of the visualization across messages from three English and Arabic forums in comparison with Support Vector Machines (SVM) and found that Writeprints provided excellent classification performance, significantly outperforming SVM in many instances. Based on our results, we believe the visualization can assist law enforcement in identifying cyber criminals and also help users authenticate fellow online members in order to deter cyber deception. © Springer-Verlag Berlin Heidelberg 2006.
Abstract:
For e-government to succeed, government agencies must manage their records and archives of which the sheer volume and diversity necessitate the use of electronic record management systems (ERMS). Using an established business-technology alignment model, we analyze an agency's strategic alignment choice and examine the outcomes and agency performance associated with that alignment. The specific research questions addressed in the study are as follows: (1) Do strategic alignment choices vary among agencies that differ in purpose or position within the overall government hierarchy? (2) Do agencies' alignment choices lead to different outcomes? and (3) Does performance in implementing, operating, and using ERMS vary among agencies that follow different alignment choices? We conducted a large-scale survey study of 3,319 government agencies in Taiwan. Our data support the propositions tested. Based on the findings, we discuss their implications for digital government research and practice.
Abstract:
Retrieving knowledge from a knowledge repository includes both the process of finding information of interest and the process of converting incoming information to a person's own knowledge. This paper explores the application of 3D interfaces in supporting the retrieval of spatial knowledge by presenting the development and the evaluation of a geo-referenced knowledge repository system. As computer screen is crowded with high volume of information available, 3D interface becomes a promising candidate to better use the screen space. A 3D interface is also more similar to the 3D terrain surface it represents than its 2D counterpart. However, almost all previous empirical studies did not find any supportive evidence for the application of 3D interface. Realizing that those studies required users to observe the 3D object from a given perspective by providing one static interface, we developed 3D interfaces with interactive animation, which allows users to control how a visual object should be displayed. The empirical study demonstrated that this is a promising approach to facilitate the spatial knowledge retrieval. © 2004 Elsevier B.V. All rights reserved.
