Please use this identifier to cite or link to this item:
Title: Domain ontology generation in web search
Authors: Lau, Steven Jia Lim.
Keywords: DRNTU::Engineering::Computer science and engineering::Information systems::Information systems applications
Issue Date: 2009
Abstract: Ontologies can be used to enhance information retrieval and interpretation greatly. They also enhance the applications readability and understandability of web documents. Ontology learning is a field in information extraction where we are concerned with semi-automatic extraction of concepts and relations from a corpus or raw data to create an ontology model. There are many methods which can achieve these aims as there are disciplines. We often require data for varied applications, many structure data differently thus the ontology model doesn’t try to model for any one application. Rather it structures the data into relevant concepts and relations, making it more meaningful and useful in other ways. The method thru this is achieved has always involved the field of natural language processing where we are concerned with the semantics and the latent relationships between each word and sentence. Algorithms such as the Hidden Markov Model (HMM) allow for learning and inference. Not only will we be able to disambiguate the latent semantics of each word but also topics within a corpus. In this report, an experiment involving a food and beverage recommendation agent mechanism is proposed where the application combines the ontology with intelligent ontology construction agent and the intelligent search mechanism agent. We shall see how the training corpus can create an HMM which can predict approximately the correct states and hence improve ontology modeling thru the 4-layer object orientated structure. An ontology model will also generate a significantly improved result based on user queries and keywords. Much more than compared to the general web search engine that we all so commonly use.
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:SCSE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
  Restricted Access
656.18 kBAdobe PDFView/Open

Page view(s) 50

checked on Sep 27, 2020


checked on Sep 27, 2020

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.