A Web Data Extraction Method For Domain Expert

Data mining is a method that it extracting implicit in the work, that people do not know in advance, but is potentially useful information and knowledge from a large number of, incompleted, noised, vagued, randomized data. Data mining technology is a new business information processing technology, mainly in the database a large number of the actual business data extraction, transformation, analysis and other processing, extracted decision support, and potentially useful information and knowledge. This knowledge mainly refers to the rules, modalities, rules, constraints, and so a broad knowledge.At present, the data mining process is mainly the work done by the data mining experts, which is due to the current data mining the characteristics of the decision itself, the data mining is a complex process, including the identification of operational issues, identify and organize data, build models and Use has been made of the information process. These require the mastery of complex data mining theories, reasonable and effective use of computer tools for processing, so data mining is mainly done by data mining experts, is inevitable during this period.The role of domain expert reflected in the domain knowledge, we know that data mining process, first of all must be domain expert of data mining engineers to acquire knowledge, through appropriate knowledge of editing software to establish the knowledge base used to guide the practice of data mining in order to gain new knowledge or rules. Expert knowledge in the field of spatial data mining process is mainly reflected in the data mining process, the implementation of the guidance, control, supervision, and verify the results of the excavation, evaluation and explanation.Domain experts in data mining process play a key role. The entire data mining process is interactive, and field-related, particularly those with a need for designers to design knowledge, experience, participation of experts, it is not a fully automated process. Term[4] stressed that the typical data mining process should be a phased process of knowledge modeling-based tests, in the establishment of any system of knowledge-based practice, in which the domain experts should play a key role.In the first phase of data mining, by domain experts and data mining analysts repeated consultations between the summing up problems, accurate definition of the original problem. In the data pre-processing phase, including data cleaning, data integration, data conversion, data reduction and other processes require the application of knowledge experts in the field to delete non-relevant attributes of the raw data abstraction, to fill the missing values, the definition of the cycle time of observation scale operations such as processing. In the data mining algorithm selection and implementation phase of mining, data mining, analysts must decide what combination of domain knowledge or a combination of technology, what technologies to deal with current problems. Select the appropriate algorithm basically a non-deterministic, and requires repeated the course of implementation. If you have expert knowledge of the guidance, will have more targeted to achieve better results.Access to the knowledge of data mining experts must be a modified and validated until the satisfaction of domain experts. Access to knowledge and interpretation of the evaluation is usually the use of a set of test data sets to measure the performance of that knowledge to evaluate the precision and accuracy of its experts in the field should also be formulated according to the relevant purpose.Domain experts more and more deeply involved in the data mining process, and even bear the main tasks of data mining will be an inevitable development of history, even though their development process may be quite long. Knowledge of experts in the field of data mining is not a data mining expert working of exclusion and negation, but rather on the data mining experts to carry out excavation work in a useful complement and steps. Human understanding of the objective world is constantly in-depth.Experts in the field an initial attempt to the database mining is the process of a gradual deepening of this understanding of the performance, only through this constant attempt to come to realize their very nature, aware of the database may be some hidden patterns, laws and knowledge. When the excavation gradually in depth, the domain experts may not be completed, when work is transferred to the data mining experts to continue to complete the data mining process. This became the initial steps of data mining and the original requirements.Domain expert mining data, will gradually increase their knowledge of data mining and the law of master and understanding, select the appropriate algorithm requires repeated practice and implementation process, will achieve better results. Interpretation of the results of data mining areas, will further enhance its right to interpret and apply the results of data mining capacity, and better used in practice.Date mining by domain expert the biggest obstacle is data mining tools and application of complex and difficult, but these barriers in the data mining experts and computer technology workers will be kept under continuous efforts to overcome and resolved. With the historical development and the progress of time, data mining technology in the mining efforts of experts, will become increasingly easy to be proficient users of other domain and use.At present, experts in the field of data mining The first difficulty is the WEB data collection problems. Many come from the data source on the Internet, how to on the Internet semi-structured data conversion as easy to deal with experts in the field of structured database is a serious problem. Some researchers had conducted a study of WEB data collection, but a solution for non-computer professionals difficult to grasp and understand, this paper presents an easy-to domain expert to understand, grasp and experts in the field of intervention to resolve the issue of WEB data collection.