Data mining could be defined as the aspect of digging out data and analyzing it with an aim of summarizing it into useful bits that could be employed in reducing costs and increasing revenues or both. In this regard, data mining remains as an important area of study that requires to be examined in detail bearing in mind the fact that it is yet to be fully embraced in the manufacturing industry which has under-utilized it as a process of finding, analyzing and utilizing the available information and data. With this respect, this research paper will study the challenges of data mining in the manufacturing industry.
Throughout the history of mankind, people have found it useful to extract past data as a way of finding out different patterns that had been used earlier on to solve different problems in the society and utilized this data to advance their lives. In line with this, there are different traditional methods of data mining that were employed in data mining. However, these trends are changing and advanced methods of data mining have so far emerged. This has resulted from the fact that there is an increase in data that is available and using traditional methods as a way of getting data and analyzing it has proved to be ineffective and time wasting (Wang, 2009, p.487).
Data mining therefore comprises of four major parts namely; clustering, classification, regression and association rule learning or rather market basket analysis. Each of these parts plays a critical role in collecting and analyzing of the available data and how this data could be processed into helpful chunks that are needed in the manufacturing industry. In reference to Witten & Frank (2005), datasets do exhibit different kinds of simple structures, some of which are relevant to the work that they are required to perform whereas some of these datasets may be completely irrelevant or redundant (p.84).
According to Wang (2007), data mining techniques remains as some of the most important tools in the manufacturing industry, yet they have often been ignored or under-utilized in most manufacturing industries (p.488). In this regard, the manufacturing industry has continued to struggle or in some cases drag due to failure in their approach to acquisition and analysis of important data. Wang (2007) asserts that there has been little research on the application of data mining techniques in the manufacturing industry (p.488). This is unlike the financial industry that has employed the techniques of data mining in collecting, analyzing and usage of important information and data.
It is important to note that interest among researchers in using data mining in the manufacturing industry begun recently. As a result of this, most researchers in the manufacturing industry are unfamiliar with the appropriate algorithm that could be employed in data mining in this sector. According to Wang (2007), there are different software companies that have developed data mining programs for non-manufacturing industry whereas there has been little focus on producing custom-based data mining software for the manufacturing industry (p.489). These therefore are some of the challenges that the manufacturing industry faces in terms of using data mining techniques in acquisition and processing of important data into information.
However, it was found out that data mining processes could go a long way in enhancing the activities of the manufacturing industry through the reduction of different costs and increase of revenues. Pham (2006) argues that different areas in the manufacturing industry stand a chance to benefit and they include productivity and quality improvements (p.665). Notably, this would require serious research on data mining specifically in relation to the manufacturing industry while at the same time working on developing appropriate data mining algorithm that is geared towards dealing with manufacturing problems.
To begin with, the processes of data mining are not as profound in the manufacturing industry as they are in any other industry such as the financial industry and other non-manufacturing industry. This is attributed to various factors. As mentioned earlier, data mining researchers have carried out little research in the manufacturing industry. Following this point, Wang (2007) argues that most researchers have limited knowledge in regard to the necessary or rather appropriate algorithm that is needed in data mining processes in the manufacturing industry. In addition, most theoretical data mining researchers are not conversant with the complex domain area of the manufacturing industry. According to Pham (2006), one cannot be able to implement any data mining process he or she has to be familiar with the data mining models and algorithms that are necessary for that particular task (p.655).
Therefore, the greatest predicament to implementing data mining in the manufacturing industry is lack of the necessary knowledge and skills that are needed for this process to be accomplished. One way that has been presented as a way of overcoming these challenges is forming a collective responsibility by bringing together researchers from different domains that would then work together and develop the necessary algorithm for the manufacturing industry. In other words, there is need for an effective collaboration among potential experts (Wang, 2007, p.494). These researchers can also examine the benefits that a particular manufacturing industry stand to gain when implementing data mining processes in the processing of its data into important information.
In summation, data mining is one of the most important processes of collecting, analyzing and utilization of data that is yet to be fully embraced in the manufacturing industry. This is as a result of the fact that there is limited knowledge on the algorithm and models that could be utilized in the manufacturing industry. Therefore, to counter this challenge, there is need for experts from different domains to collaborate and come up with appropriate models and algorithm in regard to data mining in the manufacturing industry.