--- title: "First two phases of KDD" subject: "Data Mining" chapter: 1 key: "semester_7/data_mining:kdd-first-two-phases" source: https://www.bctengineer.com/q/semester_7/data_mining/kdd-first-two-phases site: BCT Engineering --- # First two phases of KDD Data Mining · Chapter 1: Introduction - Asked in: year not given, 2+6 marks - Video: [Introduction to Data Mining and KDD Process | Data Mining Tutorial | Ep 1](https://www.youtube.com/watch?v=2NCa8hn4Kos) — Kazi Amit Hasan - Answer checked by Claude Opus 5.5 (high effort) on 2026-10-14 - Page: https://www.bctengineer.com/q/semester_7/data_mining/kdd-first-two-phases - Whole chapter: https://www.bctengineer.com/md/semester_7/data_mining/chapter-1 ## Question Is KDD a data mining? Explain the first two phases of KDD with example. ## Short answer KDD is the overall knowledge discovery process; data mining is just one step. The first two phases are data cleaning (removing noise, filling missing values) and data integration (merging multiple sources). Example: a university cleans marks and integrates exam, library, and fee databases. ## Answer **Is KDD data mining?** No. **KDD (Knowledge Discovery in Databases)** is the whole process of discovering useful knowledge from data. It has several steps: cleaning, integration, selection, transformation, data mining, pattern evaluation and presentation. **Data mining** is only one of these steps, the one in which intelligent algorithms extract patterns. So data mining is a part of KDD, although the two names are often used as synonyms. #### First phase: Data cleaning Data cleaning removes noise and inconsistent data and handles missing values, because real data is dirty. Techniques: fill missing values (mean, most probable value), smooth noise by binning or regression, remove duplicates and correct inconsistencies. #### Second phase: Data integration Data integration combines data from multiple sources (databases, files, data cubes) into one coherent store. Issues: schema integration, entity identification (cust_id vs customer_no), redundancy (detected by correlation analysis) and conflicting values (different units). #### Example: university A university wants to find the factors behind student failure. - **Cleaning**: the marks table has blanks (missing marks are replaced with the subject average) and a mark of 950 (a typo, corrected to 95); duplicate student records are removed. - **Integration**: the exam database (`roll_no`), library database (`student_id`) and fee database (`reg_no`) are combined into one student table by matching these identifiers; marks stored as percentages and as GPA are converted to one unit. The cleaned, integrated data is then selected, transformed and mined. ## Check yourself The answer's key points as multiple-choice questions, with the correct option and why. 1. What is the relationship between KDD and data mining? A. They are exactly the same process B. KDD is only one step of data mining C. Data mining is one step within the overall KDD process D. Data mining comes before KDD Answer: C. KDD is the whole discovery process; data mining is the step where algorithms extract patterns. (Key point: KDD vs data mining) 2. What is done in the first phase, data cleaning? A. Combining several sources into one store B. Removing noise and inconsistencies and handling missing values C. Converting data into mining forms like normalisation D. Showing patterns to the user Answer: B. Cleaning handles noise, inconsistent data and missing values because real data is dirty. (Key point: Data cleaning purpose) 3. Which issue belongs to data integration? A. Entity identification, e.g. cust_id vs customer_no B. Filling a blank mark with the mean C. Smoothing noise by binning D. Evaluating patterns by support Answer: A. Integration deals with schema integration, entity matching, redundancy and conflicting values from multiple sources. (Key point: Data integration issues) 4. In the university example, the exam, library and fee databases (roll_no, student_id, reg_no) are combined into one student table. Which phase is this? A. Data cleaning B. Data integration C. Data selection D. Pattern evaluation Answer: B. Matching identifiers across sources and merging them into one coherent store is data integration. (Key point: University example) ## Related in Data Mining - [Data mining as a step in KDD](https://www.bctengineer.com/question.md?key=semester_7%2Fdata_mining%3Adm-as-kdd-step) — Data Mining, chapter 1 - [KDD phases with example](https://www.bctengineer.com/question.md?key=semester_7%2Fdata_mining%3Akdd-phases-example) — Data Mining, chapter 1, asked 3× - [KDD applied to WWW mining](https://www.bctengineer.com/question.md?key=semester_7%2Fdata_mining%3Akdd-in-www-mining) — Data Mining, chapter 7 - [Data mining and its applications](https://www.bctengineer.com/question.md?key=semester_7%2Fdata_mining%3Adm-banking-biomedical) — Data Mining, chapter 1 - [Data mining vs DW and applications](https://www.bctengineer.com/question.md?key=semester_7%2Fdata_mining%3Adm-vs-dw-applications) — Data Mining, chapter 1 - [CRM; security and privacy of data mining](https://www.bctengineer.com/question.md?key=semester_7%2Fdata_mining%3Acrm-security-privacy) — Data Mining, chapter 7 ## Related in other subjects - [Data warehouse and data mining](https://www.bctengineer.com/question.md?key=semester_8%2Finformation_system%3Adata-warehouse-data-mining) — Information System, chapter 4, asked 2× - [Short note: data mining and warehousing](https://www.bctengineer.com/question.md?key=semester_8%2Finformation_system%3Adata-mining-and-warehousing-note) — Information System, chapter 4 - [Data mining and business knowledge](https://www.bctengineer.com/question.md?key=semester_8%2Finformation_system%3Adata-mining-business-knowledge) — Information System, chapter 4 - [Townsend's first and second coefficients from data](https://www.bctengineer.com/question.md?key=semester_8%2Fhigh_voltage_engineering%3Atownsend-first-second-coefficients) — High Voltage Engineering, chapter 6