Application of hermeneutics to studying an experience mining process

In this paper we investigate an approach to eliciting practitioners’ problem-solving experience across an application domain. The approach is based on a well-known ‘pattern mining’ process which commonly results in a collection of sharable and reusable ‘design patterns’. While pattern mining has been recognised to work effectively in numerous domains, its main problem is the degree of technical proficiency that few domain practitioners are prepared to master. In our approach to pattern mining, patterns are induced indirectly from designers’ experience, as determined by analysing their past projects, the problems encountered and solutions applied in problem rectification. Through the cycles of hermeneutic revisions, the pattern mining process has been refined and ultimately its deficiencies addressed. The hermeneutic method used in the study has been clearly shown in the paper and illustrated with examples drawn from the multimedia domain. The resulting approach to experience elicitation provided opportunities for active participation of multimedia practitioners in capturing and sharing their design experience.

Reducing performance bias by intrinsically insensitive learning for unbalanced text mining

This thesis proposes three effective strategies to solve the significant performance-bias problem in imbalance text mining: (1) creation of a novel inexact field learning algorithm to overcome the dual-imbalance problem; (2) introduction of the one-class classification-framework to optimize classifier-parameters, and (3) proposal of a maximal-frequent-item-set discovery approach to achieve higher accuracy and efficiency.

Mining frequent itemsets in distorted databases with granula computing

Data perturbation is a popular method to achieve privacy-preserving data mining. However, distorted databases bring enormous overheads to mining algorithms as compared to original databases. In this paper, we present the GrC-FIM algorithm to address the efficiency problem in mining frequent itemsets from distorted databases. Two measures are introduced to overcome the weakness in existing work: firstly, the concept of independent granule is introduced, and granule inference is used to distinguish between non-independent itemsets and independent itemsets. We further prove that the support counts of non-independent itemsets can be directly derived from subitemsets, so that the error-prone reconstruction process can be avoided. This could improve the efficiency of the algorithm, and bring more accurate results; secondly, through the granular-bitmap representation, the support counts can be calculated in an efficient way. The empirical results on representative synthetic and real-world databases indicate that the proposed GrC-FIM algorithm outperforms the popular EMASK algorithm in both the efficiency and the support count reconstruction accuracy.

Multiagent framework for bio-data mining

This paper proposes to apply multiagent based data mining technologies to biological data analysis. The rationale is justified from multiple perspectives with an emphasis on biological context. Followed by that, an initial multiagent based bio-data mining framework is presented. Based on the framework, we developed a prototype system to demonstrate how it helps the biologists to perform a comprehensive mining task for answering biological questions. The system offers a new way to reuse biological datasets and available data mining algorithms with ease.

A triple-random ensemble classification method for mining multi-label data

This paper presents a triple-random ensemble learning method for handling multi-label classification problems. The proposed method integrates and develops the concepts of random subspace, bagging and random k-label sets ensemble learning methods to form an approach to classify multi-label data. It applies the random subspace method to feature space, label space as well as instance space. The devised subsets selection procedure is executed iteratively. Each multi-label classifier is trained using the randomly selected subsets. At the end of the iteration, optimal parameters are selected and the ensemble MLC classifiers are constructed. The proposed method is implemented and its performance compared against that of popular multi-label classification methods. The experimental results reveal that the proposed method outperforms the examined counterparts in most occasions when tested on six small to larger multi-label datasets from different domains. This demonstrates that the developed method possesses general applicability for various multi-label classification problems.