Data mining as a cognitive tool: Capabilities and limits


Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License

Due to the large volumes of empirical digitized data, a critical challenge is to identify their hidden and unobvious patterns, enabling to gain new knowledge. To make efficient use of data mining (DM) methods, it is required to know its capabilities and limits of application as a cognitive tool. The paper aims to specify the capabilities and limits of DM methods within the methodology of scientific cognition. This will enhance the efficiency of these DM methods for experts in this field as well as for professionals in other fields who analyze empirical data. It was proposed to supplement the existing classification of cognitive levels by the level of empirical regularity (ER) or provisional hypothesis. If ER is generated using DM software algorithm, it can be called the man-machine hypothesis. Thereby, the place of DM in the classification of the levels of empirical cognition was determined. The paper drawn up the scheme illustrating the relationship between the cognitive levels, which supplements the well-known schemes of their classification, demonstrates maximum capabilities of DM methods, and also shows the possibility of a transition from practice to the scientific method through the generation of ER, and further from ER to hypotheses, and from hypotheses to the scientific method. In terms of the methodology of scientific cognition, the most critical fact was established – the limitation of any DM methods is the level of ER. As a result of applying any software developed based on DM methods, the level of cognition achieved represents the ER level.

view full abstract hide full abstract
    • Figure 1. General scheme of cognition
    • Figure 2. Correlation between thinking, reality, and sign systems
    • Figure 3. Relationship between the levels of cognition
    • Figure 4. Search scheme for hidden empirical regularities
    • Conceptualization
      Maxim Polyakov, Igor Khanin, Gennadiy Shevchenko, Vladimir Bilozubenko
    • Formal Analysis
      Maxim Polyakov, Vladimir Bilozubenko
    • Investigation
      Maxim Polyakov
    • Supervision
      Maxim Polyakov, Igor Khanin, Gennadiy Shevchenko
    • Writing – original draft
      Maxim Polyakov, Igor Khanin, Gennadiy Shevchenko, Vladimir Bilozubenko
    • Writing – review & editing
      Maxim Polyakov, Igor Khanin, Gennadiy Shevchenko, Vladimir Bilozubenko
    • Methodology
      Igor Khanin, Gennadiy Shevchenko, Vladimir Bilozubenko
    • Project administration
      Gennadiy Shevchenko