Mahidol University Logo
Faculty of ICT, Mahidol University
 

Admissions

Printable Version

 

FINDING SUBJECTIVE INTERESTINGNESS PATTERNS BY USING META-RULES GUIDED MINING BASED ON ONTOLOGY

 

TITLE FINDING SUBJECTIVE INTERESTINGNESS PATTERNS BY USING META-RULES GUIDED MINING BASED ON ONTOLOGY
AUTHOR PREECHA TANGWORAKITTHAWORN
DEGREE MASTER OF SCIENCE PROGRAMME IN COMPUTER SCIENCE
FACULTY FACULTY OF SCIENCE
ADVISOR JARERNSRI L. MITRPANONT
CO-ADVISOR THANWADEE T. SUNETNANTA
 
ABSTRACT
This research proposes a novel framework to construct the Meta-Rules Guided Mining by applying the ontology to describe the Conceptual Model of the objects in the domain of interest. The proposed framework is an “Extended Version” of traditional data mining in which the focus is on the post analysis of the data mining. The proposed framework consists of two main parts. First, the specialized rules discovery part is responsible for discovering the patterns of data from the prepared database based on the Apriori Algorithm. Second, the generalized rules discovery part is used to discover the “Generalized Rules” defined as the meta-rules patterns which are generated based on the ontology. In order to represent the categorical concepts and relation types of the ontology, the Formal Concept Analysis (FCA) theory was adopted to construct the lattice. The prototype system called FCA2MR (Formal Concept Analysis to Meta- Rules) was developed as an easy-to-use and step-by-step process generation tool. It was developed to test and verify the idea of the proposed approach. The FCA2MR provides a flexible way to specify the pre-threshold measurement including minimum support and minimum confidence values to discover interestingness patterns. In addition, the postthreshold measurement called minimum precision values was used to determine the strength of the generated meta-rules patterns. In the experiment, three benchmark datasets from the UCI machine learning repository were used to test the prototype and three active datasets were collected and used to evaluate the interestingness results. The system was evaluated in three parts. First, the functional comparison of the ontology generators was conducted to compare the ontology features of three ontology generators, namely, FCA2MR, Prot้g้2000 and ConImp. Second, the interestingness evaluation showed comparable results between the FCA2MR and the Magnum Opus program. Finally, the user participation evaluation was carried out by using a number of questionnaires to check whether the ranking results reflect the subjective interestingness of the users. The experimental results indicated that this approach is feasible and effective in generating reasonable solutions for subjective interestingness data mining comparable to commercial tools. Moreover, the prototype system gives high accuracy and the interestingness results are relevant to the user’s needs.
KEYWORD DATA MINING/ INTERESTINGNESS/ SUBJECTIVE INTERESTINGNESS/ META-RULES GUIDED MINING/ ONTOLOGY/ FORMAL CONCEPT ANALYSIS (FCA)

 

Go to Top

 

ICT Building, Mahidol University, 999 Phuttamonthon 4 Road, Salaya, Nakhonpathom 73170 Tel. +66 02 441-0909 Fax. +66 02 849-6099
Mahidol University Computing Center, The Faculty of ICT, Mahidol University , Rama 6 Road, Rajathevi, Bangkok 10400 Tel. +66 02 354-4333 Fax. +66 02 354-7333