Peerasak Intarapaiboon

A hierarchy-based similarity measure for intuitionistic fuzzy sets

Due to some unreasonable results obtained from most current similarity measures for intuitionistic fuzzy sets (IFSs), we introduce a necessary condition to obtain a stronger definition of similarity measures for IFSs, and present a new similarity measure derived from a general idea of similarity measures for concepts on a lattice. In experiments, we focus our attention on two basic directions of performance evaluation: one is how much the proposed measure is reasonable and the other is how much accuracy the measure produces when it is applied to classification problems. The experimental results show that the proposed measure is reasonable and achieves a satisfactory performance on classification problems.

Information Extraction from Thai Text with Unknown Phrase Boundaries

Using sliding-window rule application and extraction filtering techniques, we propose a framework for extracting semantic frames from Thai textual phrases with unknown boundaries based on patterns of triggering terms. A supervised rule learning algorithm is used for constructing multi-slot extraction rules from hand-tagged training phrases. A filtering module is introduced for predicting rule application across phrase boundaries based on instantiation features of rule internal wildcards. The framework is applied to text documents in three domains with different target-phrase density and average lengths. The experimental results show that the filtering module improves precision and preserves high recall satisfactorily, yielding extraction performance comparable to frame extraction with manually identified phrase boundaries.

Extracting Semantic Frames from Thai Medical-Symptom Phrases with Unknown Boundaries

Due to the limitations of language-processing tools for the Thai language, pattern-based information extraction from Thai documents requires supplementary techniques. Based on sliding-window rule application and extraction filtering, we present a framework for extracting semantic information from medical-symptom phrases with unknown boundaries in Thai free-text information entries. A supervised rule learning algorithm is employed for automatic construction of information extraction rules from hand-tagged training symptom phrases. Two filtering components are introduced: one uses a classification model for predicting rule application across a symptom-phrase boundary, the other uses extraction distances observed during rule learning for resolving conflicts arising from overlapping-frame extractions. In our experimental study, we focus our attention on two basic types of symptom phrasal descriptions: one is concerned with abnormal characteristics of some observable entities and the other with human-body locations at which symptoms appear. The experimental results show that the filtering components improve precision while preserving recall satisfactorily.

An Improvement of Unknown-Item Search for OPAC Using Ontology and Academic Information

Many students usually use the unknown-item search strategies, including subject and keyword searches, to retrieve books or other materials provided in library catalogs. However, the success rates for unknown-item searching is relatively low comparing with the known-item search strategies, i.e., title or author searches. In this paper, a framework for improving the unknown-item search is proposed. The main contributions of our framework are concerned with both user’s keywords and book indexing: (i) To enhance a user’s keyword, the framework will select other relevant terms in a domain-related ontology. (ii) Topics expressed in course description are used as book indexing. A preliminary experiment shows that the traditional OPAC incorporating with the proposed framework gives satisfactory results.

An Improvement of Pattern-Based Information Extraction Using Intuitionistic Fuzzy Sets

Multi-slot information extraction (IE) is a task that identify several related entities simultaneously. Most researches on this task are concerned with applying IE patterns (rules) to extract related entities from unstructured documents. An important obstacle for the success in this task is unknowingness where text portions containing interested information are. This problem is more complicated when involving languages with sentence boundary ambiguity, e.g. the Thai language. Applying IE rules to all reasonable text portions can degrade the effect of the obstacle, but it raises another problem that is incorrect (unwanted) extractions. This paper aims to present a method for removing incorrect extractions. In the method, extractions are represented as intuitionistic fuzzy sets (IFSs), and a similarity measure for IFSs is used to calculate distance between IFS of an unclassified extraction and that of each already-classified extraction. The concept of k nearest neighbor is adopted to design whether the unclassified extraction is correct of not. From the preliminary experiment on a medical domain, the proposed technique improves extraction precision while satisfactorily preserving recall.

An Improvement of IFS-Based Classification Using Correlation Coefficient between Features

Over the past decades, theoretical and application researches of similarity measures between intuitionistic fuzzy sets (IFSs) have been continuously revealed. Solving pattern classification problems is one of most prominent areas to which these similarity measures can be applied. Differing from other aspect frameworks for classification, IFS-based frameworks do not take relationship among features into account. In the present paper, a modified IFS-based framework by using correlation coefficient among features is presented. The experimental results on various real-world problems show that the proposed framework achieves a satisfactory performance.

A Framework for Text Classification Using Intuitionistic Fuzzy Sets

Due to massively increasing of web pages and online documents, one of crucial processes to handle those documents is automatic (or at least semi-automatic) text classification. Based on the concept of intuitionistic fuzzy set (IFS), a framework for text classification is presented. In the framework, we introduce statistical methods to represent each document as an IFS and to learn a pattern of each document class. Then, a similarity measure for IFSs is applied in order to assign the most relevant class to a new document. The proposed framework with various similarity measures for IFSs is evaluated by benchmark datasets. The experimental results show that our framework yields satisfactory results.

Extracting chemical reactions from Thai text for semantics-based information retrieval

Based on sliding-window rule application and extraction filtering, we present a framework for extracting multi-slot frames describing chemical reactions from Thai free text with unknown target-phrase boundaries. A supervised rule learning algorithm is employed for automatic construction of pattern-based extraction rules from hand-tagged training phrases. A filtering method is devised for removal of incorrect extraction results based on features observed from text portions appearing between adjacent slot fillers in source documents. Extracted reaction frames are represented as concept expressions in description logics and are used as metadata for document indexing. A document knowledge base supporting semantics-based information retrieval is constructed by integrating document metadata with domain-specific ontologies.