Abstract
In this paper, we present a new technique which is the Admixture MCRDR-FCA (AMF) algorithm for Web document classification. The technique offers a practical approach to new Web document classification by combining and extending a number of the current techniques. The AMF algorithm has a number of noteworthy features: firstly, it provides a structured conceptual correlation between keywords and secondly it is optimised. Finally, the algorithm creates refined multiple new rules in order to achieve higher accuracy in the conclusions relating to document classification. This is achieved by clarifying the relationships between one concept and another concept before going on to provide a final classification to some category. To evaluate the AMF algorithm, we have developed a demonstration system that permits easy comparison with a number of other classification techniques.
| Original language | English |
|---|---|
| Title of host publication | Full Proceedings of the 2nd International Conference on Information Management and Business (IMB 2006) |
| Editors | Bhuvan Unhelkar, Yi-Chen Lan |
| Place of Publication | Sydney |
| Publisher | University of Western Sydney |
| Pages | 562-573 |
| Number of pages | 12 |
| ISBN (Print) | 174108122X |
| Publication status | Published - 2006 |
| Event | International Conference on Information Management and Business (2nd : 2006) - Sydney Duration: 13 Feb 2006 → 16 Feb 2006 |
Conference
| Conference | International Conference on Information Management and Business (2nd : 2006) |
|---|---|
| City | Sydney |
| Period | 13/02/06 → 16/02/06 |
Keywords
- ontologies
- knowledge acquisition
- web document classification
- multiple classification ripple down rules
- formal concept analysis