DATA MINING

 Data mining is a critical step in data analeptics. In layman's language, we can look at data mining as a person digging through a cave in search of gold. The method of analyzing data from various viewpoints using various methods and summarizing it into valuable information is known as data mining (Roiger, 2017.). Data mining with proper tools may result in discovering previously undiscovered fascinating patterns, unexpected records, and relationships (Witten, et al.,2005). 

figure: 1 TYPES OF DATA 

Types of data: 

In the above figure: 1 we can see that data is classified into qualitative and quantitative, which further is classified into nominal & ordinal data and discrete & continuous data.

Qualitative data: This category of data is the kind of classification that is used to group data—for example, age, gender, and one's job title. Within the quantitative data, there are nominal data and ordinal data.

·       Nominal data – data types that are not measurable. The result for this data is usually like yes/no or true/ false.

·      Ordinal data - measurable data types. The results for this data can be categorized as a low, medium, and high or average, good, and excellent.

Quantitative data: Numerical data includes data that is made up of numbers. E.g., income, distance, and temperature. Within the quantitative data, there are discrete data and continuous data.

·       Discrete data – data that is finite and whole numbers, for instance, no. of students in a class

·      Continuous data – data is infinite and is usually in a range of numbers, for instance, temperature.

 

If data mining is carried out correctly, many positive benefits could be realized. Some of these benefits include the following:

Þ    the identification of previously unrecognized connections between business data sets

Þ    good forecasting of future trends and behaviors

Þ    the extraction of value from large volumes of data (Lei-da Chen, and Frolick, 2000.)

Þ    the creation of business actions built on data insights

For instance, observing consumer behavior and purchase patterns can tell us how well our product is attractive to consumers, how likely they are to repurchase it, and, if so, how often (Lisnawati, and Sinaga, 2020.).

figure: 2  STEPS IN DATA MINING

Steps in data mining: 

As shown in the figure: 2 there are six consecutive steps in data mining.

Defining your company's objectives: It is essential to clearly understand your business's objective. To accurately specify the project's parameters, you must decide on the critical data collection goals.

Understanding data sources: with the project specifications, you will be able to determine which platforms are reliable for data collection.

Preparing data: This will prepare your data, ensuring it comes from a wide range of well-chosen sources. Data is retrieved from the source systems, processed, and uploaded to the final system.

Analyzing data: In this phase, the compiled data is sent into a sophisticated program where various machine algorithms work towards connections and patterns that may be used to guide analysis, make predictions, and inform action. Data points, also known as bits of data, are sorted, and their connections are synchronized using this program.

Studying results: You can decide if and how successfully the model's findings and insights can aid in validating your forecasts, answering your queries, and ultimately archiving the business aim.

Implementation or Deployment: After the data mining process, the findings must be communicated to decision-makers through a report. Then, they might decide how to utilize this knowledge to achieve the business purpose. Here, the system analysis is transferred to the actual world.

 

Examples of successful implementation of data mining:

According to (tutor2u, 2018) Tesco, the major supermarket chain in the United Kingdom, with its Clubcard; gathers personal information on their customers, such as their names, addresses, shopping patterns, and product preferences, in order to analyze consumer behavior and develop marketing strategies. Through data mining, Walmart realized that "strawberry pop tarts," a simple product that is simple to heat and eat on the move, must be put near the checkout before any storm warning to attract customers and improve sales during hurricanes.

People mistakenly believe that more data equals more knowledge. However, less data, what you do with it, and how you process it determines your comprehension degree (Jackson, 2002.). Data mining may be used against you by generating erroneous insights and forecasts. However, when done correctly with the proper tools, data mining helps you to filter through chaotic data noise to determine what is relevant and then utilize this knowledge in your decision-making processes.

 

References

byjus (2022) types of data. Available at: https://byjus.com/maths/types-of-data-in-statistics/ (Accessed: 23/November/2022).

Jackson, J., (2002). Data mining; a conceptual overview. Communications of the Association for Information Systems, 8(1), p.19.

Lei-da Chen, T.S. and Frolick, M.N., (2000). Data mining methods, applications, and tools. Information systems management, 17(1), pp.67-68.

Lisnawati, H. and Sinaga, A., (2020). Data Mining with Associated Methods to Predict Consumer Purchasing Patterns. International Journal of Modern Education & Computer Science, 12(5).

Roiger, R.J., (2017). Data mining: a tutorial-based primer. Chapman and Hall/CRC.

Tutor2u (2018) data mining (introduction for business students). 2018. Available at: https://www.youtube.com/watch?v=o5ATkQuBTtw (accessed: 23/November/2022).

Witten, I.H., Frank, E., Hall, M.A., Pal, C.J. and DATA, M.,(2005), June. Practical machine learning tools and techniques. In Data Mining (Vol. 2, No. 4).