Browsing by Author "Li, KL"

Now showing 1 - 2 of 2

A unified, adjustable, and extractable biological data mining-broker
(2003)
Ho, MH
;
Chang, YS
;
Cheng, MC
;
Li, KL
;
Yuan, SM
The document formats of biological data sources typically are more versatile and more complicated than the traditional data sources. It is hard to efficiently retrieve useful information from biological data sources by traditional information retrieval technologies. In this paper, we propose a unified, adjustable, and extractable Biological Data Mining-Broker mechanism. Based on XML methodology, the mechanism provides a federated forum model to overcome the heterogeneities of both form and meaning from those different diverse biological data sources. Furthermore, the mechanism also utilizes the feedback-based direct raw and meaningful extracted cache technique to improve the efficiency and accuracy of the system. The experimental results show that our proposed system has good performance, and it is a good choice for biological data mining process with multiple heterogeneous data sources, different mining applications, and knowledge analysts. It is highly useful for target discovery and bioinformatics research projects.
The Federated Forums based Integrated Biological Data Mining Broker
(2005-06-01)
Ho, MH
;
Chang, YS
;
Li, KL
;
Yuan, SM
;
資訊工程學系
The inherent diversity and complexity of biological data sources make traditional information retrieval technologies unsuitable for handling the problems that arise with the retrieval of biological documents on the Internet. Furthermore, biological data sources in all their variety have specific domains and objectives that differ from other user interface, query formats. result sets and database organizations. To meet the problems of unification in both their syntactic and semantic aspects, we here present a unified, adjustable, and extractable Biological Data Mining-Broker (BDMB). Based on XML technology, the broker provides a federated forum model that handles heterogeneity of sources. It also uses a feedback-based utility for raw and meaningful extracted cache techniques that makes the system more efficient and accurate. The experimental results show that the system performs well and is ideal for biological data mining processes with many different data sources, mining applications, and knowledge analysts. It is very useful for target discovery and bioinformatics research projects. (c) 2004 Elsevier B.V. All rights reserved.