Large amounts of information are generated every day and recorded in various applications or devices. This information comes from the entry/exit to work to the number of cars that pass by per minute on a given road. Even all the operations that are recorded through our cell phones, computers, commercial transactions. This information not only comes from our daily lives, but the expansion of companies and the need to see the history of their operations means that increasingly large amounts of data are generated and stored. All this phenomenon has been increasing thanks to technological advances in data storage (which now has a tendency to be in the cloud), the intercommunication of equipment, the cheapening of technology and the expansion of the Internet. Data mining.
The massive generation of this data creates a problem called infoxication, having so much information that sometimes makes it impossible to organize it effectively. This is where data mining becomes useful to get the most out of the large amounts of information we have.
Link infoxication: https://www.citiuschool.com/blog/content-curator-terror-de-la-infoxicacion-marketing-digital/
Data mining is a search process performed on large data sets. With the objective of discovering patterns, trends, groupings, behaviors that can support us in decision making. It seeks to extract knowledge from the data in order to predict what can happen in different scenarios and find information that is not visible to the naked eye, using statistics and probability of information that is hidden in the data.
Data Mining Steps
Within the data mining process are the following steps:
- Determination of Objectives and Data Collection. Depending on the client’s needs, the type of information desired is asked.
- Data processing. Data are selected, cleaned and transformed.
- Model selection. A model or algorithm is created to obtain the best possible result. Algorithms developed in different areas of artificial intelligence such as: linear regression, neural networks, machine learning, etc. can be used. A first draft of the graphical visualization is also made.
- Analysis of results. It is verified if the results are coherent, logical and respond to the questions asked in the objectives.
- Updating of the model. This is done so that it does not become obsolete over time.
Advantages of using data mining
Some of the advantages of using data mining are listed below:
- It helps to manage information in a more efficient way.
- Discovering information that was hidden and we did not expect to obtain.
- Give companies the ability to offer products and services that customers need.
- Contributes to strategic decision making.
- Preventing or predicting adverse situations using models and statistical data
- The results are easy to understand and interpret.
Disadvantages of data mining
The characteristics of data mining mean that we have more advantages than disadvantages, but if there were a disadvantage to mention it would be the complexity of using different data mining techniques depending on the data you want to work with. Another would be the initial investment and this depends on what data you want to collect and if these are already in a database or it is necessary to obtain different technologies for that collection.
A high initial investment depending on the infrastructure of each company would seem a disadvantage in the implementation, but it does not seem so much considering that with data mining we can make an early detection of problems, predict asset wear and anticipate maintenance, maintain the production line in a more continuous way.
Difference between data analysis and data miningItis very easy to relate these 2 terms and in several texts they are confused or treated in the same way, but there is a thin line that separates data analysis from data mining. Both are subdomains of data science and use similar processes, however data analytics performs data processing and modeling to find useful information that suggests conclusions to support decision making by working with specific metrics and indicators. In addition, it usually does not work with a hypothesis to guide the work, but rather processes data to constantly monitor and create reports. Data mining is a data analysis technique but used with a different approach. It is used for the discovery of hidden information and prediction of behavior. Mining is more focused to be used on large data sets called Big Data.