Issue No. 07 - July (2001 vol. 34)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/2.933505
<p>Traditionally, the term community refers to a location where people with common interests gather to share experiences, ask questions, or collaborate. Because they are present in the same locale, members can meet easily to learn from each other by sharing their explicit knowledge and revealing information about their successes and failures. E-communities can use Web technologies to provide geographically disparate groups with the same sense of community. To thrive, these virtual communities require software agents that perform many of the functions that print resources and membership services provide to traditional communities. </p> <p>The authors' studies have shown that intelligent software agents, in the role of personal agents, can effectively provide data sharing, personalized services, and pooled knowledge while maintaining user privacy and promoting interaction in e-communities. Future research should include further investigation of several topics, including automatic learning of short- and long-term user interests, improving information retrieval accuracy, and ontology management. Successful deployment of multi-IDIoMS systems requires addressing ontology management--an issue that generally plagues the Internet. However, recognizing and efficiently collating information about similar topics within heterogeneous data sources remain difficult issues to resolve. </p> <p>Using XML and other standards can mitigate some concerns, but it does not provide a solution to other key issues, especially with regard to legacy data. Nonetheless, the authors believe that their agent-based information management system provides a powerful application for use in complex, distributed Internet e-communities. </p>
T. Ohtani, S. Case, N. Azarmi and M. Thint, "Enhancing E-Communities with Agent-Based Systems," in Computer, vol. 34, no. , pp. 64-69, 2001.