Data mining between classical and modern applications : a review

• Published in: AL-Rafidain journal of computer sciences and mathematics Vol. 15; no. 2; pp. 171 - 191
• Main Author: Yasin, Ammar Th
• Format: Journal Article
• Language: English
• Published: الموصل، العراق جامعة الموصل، كلية علوم الحاسبات و الرياضيات 03.12.2021; Mosul University
• Subjects: big data; data mining; knowledge discovery in databases; البيانات الضخمة; التعلم الآلي; التنقيب في البيانات; الذكاء الاصطناعي; برامج الحاسوب
• ISSN: 1815-4816; 2311-7990; 2311-7990
• DOI: 10.33899/csmj.2021.170020

Data mining (DM) is an incredible innovation with extraordinary potential to help organizations centre around the main data in the information they have gathered about the conduct of their clients and likely clients. It finds data inside the information that inquiries and reports can't viably uncover. Overall, DM (to a great extent called information or data revelation) is the route toward analysing data according to substitute perspectives and summarizing it into significant information - information that can be used to assemble pay, diminishes costs, or both. DM writing computer programs is one of different logical gadgets for separating data. It grants customers to separate data from a wide scope of estimations or focuses, organize it, and summarize the associations perceived. In reality, DM is the path toward finding associations or models among numerous fields in enormous social datasets. Procedures used in DM measure come from a mix of computational strategies including Artificial Intelligence (AI), Statistics, Machine Learning (ML), and Database (DB) Systems. Aside from the centre techniques used to do the investigation, the cycle of DM can include different pre-handling ventures preceding executing the mining method. Also, a post-preparing stage is normally utilized to picture the aftereffects of the investigation (for example perceived examples or recovered data) in an instinctive and simple to-impart way. From a wide perspective, there are two significant standards of methods: expectation and information disclosure. It includes four sub-groups: a) Classification, Prediction and Regression, b) Clustering, c) Association Rule and Sequence Pattern Mining, and d) Outliers and Anomaly Detection. What's more, there are some generally new and energizing zones of information investigation, for example, spatial DM and graph DM that have been made conceivable through the structure squares of DM techniques. This survey not just advantages analyst to create solid examination subjects and distinguish gaps in the research areas yet additionally helps experts for data mining and Big Data (BD) software framework advancement.