Knowledge discovery from databases (KDD):

Knowledge discovery from databases (KDD):

Enterprises collect and store more and more current, detailed and accurate production data. Data serves as an enormous potential and source of knowledge, huge amount of data and its complexity pose a challenge to reduce and analyze the same without the use of complex automated analysis techniques.

All manufacturing industries use powerful data acquisition systems to collect, analyze and transfer data from almost all the processes of the organization. This data may be related to machines, products, processes, maintenance, quality control, failure detection etc is stored in the data bases at various stages. The use of databases and statistical techniques are well established since it enables the industry with valuable insights and resources. Earlier they were used in order to find patterns in manufacturing operations. But with growing complexities in the manufacturing processes very large amount of data is acquired which contains hundreds of attributes / variables. In order to model the system behavior accurately, these attributes / variables have to be analyzed simultaneously. This complexity of attributes / variables calls for new techniques and tools for the automated extraction of the useful knowledge from enormous of raw data. Data mining deals with solving these problems by applying mathematical models to automatically discover patterns in the data which is already present in the data bases.

Knowledge Discovery in Databases is one step in data mining, which denotes the entire process of turning the large raw data into useful knowledge. This method includes the following important stages:

• The first step involves the understanding of the application domain, which is of great importance when analyzing manufacturing data. Successfully generation of new knowledge calls for a close collaboration of domain experts, data experts and data mining experts. All the goals and tasks of the data mining process have to be determined and all the factors, which might affect the manufacturing process, should be revealed and understood.

• Second step includes choosing the target data set. Since data mining enables only to uncover and show the patterns already present in the data, the target data set must be large enough to contain these patterns while remain concise enough to be minded in an acceptable time frame. The data sets are usually stored in various data bases which need additional integration.

• The data sets have to be pre-processed, handling missing data and removing noise.

• Final step is data mining for the extraction of patterns from the data. This involves the selection and application of appropriate data mining algorithms as well as the development of a model which describes the pattern. This is followed by the last step where the extracted patterns have to be interpreted and verified.

Here the evaluation of from the domain experts is essential to interpret the patterns into knowledge. Here the knowledge discovery from databases (KDD) steps need to be iterated several times to finally obtain the result.

References

• Data Mining for Manufacturing: Preventive Maintenance, Failure Prediction, Quality Control Andre BERGMANN Salzgitter Mannesmann Forschung GmbH; Duisburg, Germany

• U.M. Fayyad, G. Piatetsky Shapiro, and P. Smyth, ‘From Data Mining To Knowledge Discovery: An Overview. In: Advances In Knowledge Discovery And Data Mining, eds. U.M. Fayyad, G. Piatetsky-Shapiro, P. Smyth, and R. Uthurusamy, AAAI Press/The MIT Press, Menlo Park, CA., pp. 1-34, 1996.

•  W.J. Frawley, G. Piatetsky Shapiro, and C. Matheus, ‘Knowledge Discovery In Databases: An Overview. In: Knowledge Discovery In Databases, eds. G. Piatetsky-Shapiro, and W. J. Frawley, AAAI Press/MIT Press, Cambridge, MA., pp. 1-30, 1991.

• M.H. Lee, ‘Knowledge Based Factory’, Artif. Intell. Eng., 8, pp109-125,1993.