A hybrid system for concept-based web usage mining
International Journal of Hybrid Intelligent Systems
The needs and benefits of Text Mining applications on Post-Project Reviews
Computers in Industry
Identifying fall-related injuries: Text mining the electronic medical record
Information Technology and Management
Virtual Communities of Practice's Purpose Evolution Analysis Using a Concept-Based Mining Approach
KES '09 Proceedings of the 13th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems: Part II
Expert Systems with Applications: An International Journal
Conceptual classification to improve a web site content
IDEAL'06 Proceedings of the 7th international conference on Intelligent Data Engineering and Automated Learning
| Hi-index | 0.00 |
This paper presents a Text Mining approach for discovering knowledge in texts to later construct decision support systems. Text mining can take advantage of knowledge stored in textual documents, reducing the effort for knowledge acquisition. The approach consists in performing a mining process on concepts present in texts instead of working with words. The assumption is that concepts represent real world events and characteristics better than words, allowing the understanding and the explanation of the reasoning used in decision processes. The proposed approach extracts concepts expressed in natural phrases, and then analyzes their distributions and associations. Concepts distributions and associations are used to characterize classes or situations. After the discovery process, the obtained knowledge can be embedded in automated systems to classify elements or to suggest actions or solutions to problems. In this paper, experiments using the approach in a psychiatric domain are discussed. Concepts extracted from textual medical records represent patients' symptoms, signals and social/behavior characteristics. An automatic system was constructed with the approach: a classifier whose goal is to help physicians in disease diagnoses. Results from this system show that the approach is feasible for constructing decision support systems with satisfactory performance.