Dissertation > Excellent graduate degree dissertation topics show

Research of Text Mining about Semantic Relation Recognition

Author: LiuMingYan
Tutor: YanDuanWu；ZhangChengZhi
School: Nanjing University of Technology and Engineering
Course: Management Science and Engineering
Keywords: Semantic relationships found Text Mining Body automatically build Conceptual system
CLC: TP391.1
Type: Master's thesis
Year: 2010
Downloads: 382
Quote: 1
Read: Download Dissertation

Abstract

Text Mining (Text Mining), also known as text data mining or knowledge discovery from text databases is obtained from unstructured text information the user is interested or useful model. Which is generally recognized text mining is defined as follows: text mining is the process of extracting previously unknown, comprehensible, and eventually used knowledge from large amounts of text data, and apply this knowledge to better organize information for future reference. Semantic relations found in text mining is the current research focus, the main idea is scanning and automated processing of natural language text, found that the semantic relationships between the concept of the term and concept of the term. Semantic relations between concepts is an important manifestation of knowledge, these semantic relationships have the upper / lower (hypernymy / hyponymy), part of the overall relationship (part-whole), causality (causality) synonymous relationship ( synonymy), antisense relations (antonymy) and inference relationship (inference). From a theoretical perspective, the Semantic relations found in text mining research will enable natural language processing from lexical analysis, parsing level of depth to the level of semantic analysis; From an application perspective, the Semantic relations found in text mining research can automatic or semi-automatic ontology building the basis of the theory and methods. Corpus in the field of military aircraft to the semantic relationship between the concept of the field of military aircraft system for the study, the use of text mining processing ideas and basic processes, combined with natural language processing, information extraction, ontology automatically build research theory and semantic relationships based on text mining research and exploration, and to carry out the experiment. The main work and research contents include the following four areas: (1) Semantic relations found in text mining theory and Research. Introduced the research progress both at home and abroad on natural language processing, text mining, automated build body and review is made on the direction of research and research objectives of this article. On this basis, the research content and research methods specifically addressed. (2) construction of the corpus of text processing in the field of military aircraft. Articles in Wikipedia and CNKI database related to military aircraft, as the corpus of the text corpus. The corpus of the military aircraft including the 1951 term, 304 articles, in which 3324 phrases extracted. The experiments provide the foundation for the construction of the corpus-based research, but also for subsequent automatic body building or other work material and research support. (3) research and experimentation discovered semantic relationships based on template matching. According to the structural characteristics of knowledge in the field of military aircraft, the proposed the typical semantic relationships contained in the field of military aircraft conceptual system. To automatically discover and identify these semantic relationships, combine the edit distance template matching method. The first artificial participation train a corpus, to obtain the corresponding relationship between the template and the semantic relationship type. The edit distance relationship template summarized matching, and subsequently used in the test corpus, in order to verify the efficacy of the method. (4) Based on the complex network of semantic relations research and experimentation. Combing the complex network theory of knowledge, the use of the network natural language features, the use of complex networks to discover the semantic relations. The proposed use of complex networks and related analytical tools to aid discovery field concept terms and their semantic relationships. The term and the associated word as a network node, construct a complex network, each community thus formed represents a relationship. Terminology as nodes found relationship as a side, constructed military aircraft conceptual system in the field of complex networks, and its analysis.

Related Dissertations

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Text Processing
© 2012 www.DissertationTopic.Net Mobile