| Nowadays, the world is in the period of informationization and internetization. with the rapid progress of high technologies, such as modern communications, network technology, storage technology, computer technology and multimedia applications, the digita library has also accelerated the pace of the process. The digital library is not only the trend of the development of internetization and a nevessary codition of achieving the sharing of resources, but also the vital symbol of information storage and the level of retrieval. America was the first to study digital library in 1991, then, the United Kingdom, France, Japan, Germany, Italy and other countries have also started to invest huge sums of money to develop its own digital library. China's digital library was built in 1997 and in the same year, the State Development Planning Commission approved the project ----Test-type based digital library.In ten years, digital libraries have developed rapidly in China, whcih has changed the traditional static library into a multi-media access, long-distance transmission, intelligent search, seamless cross-link library, therefore, it creates a super dimension information. However, due to large amount of information and traditional way of restricting access, the users often detect a large number of irrelevant information in complicated information. To get wanted information is laborious and time-consuming.I am a Library Reference Department staff, and often guide the users to retrieve information or do it on behalf of the users. Therefore, I am deeply impressed by the changes of digital information services. Paper-based documents service has changed gradually to the digital information services, nonetheless, it is very difficult for the users to get wanted information in a great amout of information sourse. Especially for general agricultural workers, when they encountere technical problems, they find difficult to get solutions to problems. If there is a convenient and easy way for the agricultural workers to get information without learning technology ,it will greatly improve the utilization of the inquiry. Based on this idea, I envision a set of ontology-based semantic query system of crop pests and diseases which will introduce the commone symptoms of the diseases and insect pests and solutions to the database. In order to facilitate the retrieval, I try to introduce ontology theory to the construction of the database.At present, the hot topic in the field of information retrieval is semantic search based on the concept(Ontology Retrieval ). Integrating Ontology into the traditional information, it can not only inherit adcantages of concept information retrieval, but also overcome its limitations. Since Ontology has been made it has been concerened by the number of foreign Researchers. However, The domestic-related research and application have been carried out in recent years. It is sure to bring a revolution to the field of information retrieval based on ontology even though the research is in the beginning part and there is no uniform standard.The task of the dissertation it to study the key technoligies in the field of semantic retrieval, including the construction of the field of ontology; Semantic similarity calculation model; Correlation calculation model and the development of Semantic Retrieval System. This dissertation originally designed to develop a Ontology database of crop pests and diseases. Time is limited, it just taks crop as an example, it build a corn pest control ontology based on full study of ontology and the basis of technology. In the process of constructing, it used the seven-step theory given by Noyhe Hafner In Stanford University. That is, an initial framework of ontology is given first , then add details in revising and refining the process. Based on this theory and the form of three groups(concept; Properties;value), the ontology database can be constructed. Then, analyze, compare and improve the semantic similarity model, then a new model for calculating the semantic similarity NYCX can be made after analyzing, comparing, improving semantic similarity NYCX. In constructing ontology, it used standard OWL recommended by the W3C to describe ontology. In selecting tools, it used Protege 3.3 which is easy to use and with a good graphical interface, and supporting a variety of body language to build.Next, it will make a study of the semantic similarity calculation model and semantic correlation calculation model and the two concepts are easy to confuse: the former reflects the level of two concepts while the latter shows a dependent elationship between the concepts. The concept of the semantic similarity and relevance of the calculation are two key technologies in semantic retrieval. Good computing model can provide recall and precision to a large degree. The concept of the similarity calculation can solve the concept of similarity which is difficult to quantify the problem in the traditional knowledge. NYCX computing model takes the concept of hierarchy and attributes bound into account.When measured against existing model, it gives more presice description of semantic distance between concepts.It designed a retrieval system Of corn pest control based on NYCX in computing model which includes Crop name, inquiry of disease type and inquiry of crop pests pictures. This is a module by using SPARQL to query language and support RDQL. So that the query engine associated with the relational database, which achievs higher efficiency in querying related database .This database has stored more than 150 species of plant diseases and insect pests of corn and more than 500 related articles. Data from the "Chinese Journal Network" and "Million database" and a number of agriculture-related sites, therefore, it has certain authority and relevance.The software research based on Ontology information retrieval model, it seeks to establish a unified standard interface for information retrieval systems which stresses the compatibility, scalability, availability, reliability. The system is so simple in interface that the users can easily enter and modify the input information. In the search , it is a good try that we seek to enhance the supporting of semantics and improve the qualities of Retrieval System. However, due to time and many other reasons, the contents of this paper is still at a basic stage there is still much work to be done in the future. |
PDF Full Text Request