A Systematic Review of Learning based Notion Change Acceptance Strategies for Incremental Mining
The data generated contemporarily from different communication environments is dynamic in content different from the earlier static data environments. The high speed streams have huge digital data transmitted with rapid context changes unlike static environments where the data is mostly stationery. The process of extracting, classifying, and exploring relevant information from enormous flowing and high speed varying streaming data has several inapplicable issues when static data based strategies are applied. The learning strategies of static data are based on observable and established notion changes for exploring the data whereas in high speed data streams there are no fixed rules or drift strategies existing beforehand and the classification mechanisms have to develop their own learning schemes in terms of the notion changes and Notion Change Acceptance by changing the existing notion, or substituting the existing notion, or creating new notions with evaluation in the classification process in terms of the previous, existing, and the newer incoming notions. The research in this field has devised numerous data stream mining strategies for determining, predicting, and establishing the notion changes in the process of exploring and accurately predicting the next notion change occurrences in Notion Change. In this context of feasible relevant better knowledge discovery in this paper we have given an illustration with nomenclature of various contemporarily affirmed models of benchmark in data stream mining for adapting the Notion Change.
