- Year not given · 2+8 marks
KDD applied to WWW mining
Short answer
KDD is the overall process of discovering useful knowledge, with steps cleaning, integration, selection, transformation, mining, evaluation and presentation. In web mining, selection means crawling pages and collecting logs, cleaning removes tags and noise, mining applies content, structure and usage algorithms, and evaluation filters trivial patterns.
What is KDD? How do you apply KDD in World Wide Web mining?
KDD (Knowledge Discovery in Databases) is the overall process of discovering useful, valid, previously unknown knowledge from large data. Data mining is the central step of this process.
KDD steps
- Data cleaning: remove noise and inconsistent data.
- Data integration: combine multiple sources.
- Data selection: choose data relevant to the task.
- Data transformation: normalise or aggregate into forms suitable for mining.
- Data mining: apply algorithms to extract patterns.
- Pattern evaluation: identify the interesting patterns by measures.
- Knowledge presentation: visualise and report the knowledge.
Applying KDD to World Wide Web mining
| KDD step | In web mining |
|---|---|
| Selection | Resource finding: crawl and retrieve web pages, collect server logs |
| Cleaning | Remove HTML tags, stop words, image requests, robot accesses and failed requests |
| Integration | Combine page content, link structure, logs, user registration data |
| Transformation | Convert to features: term vectors (TF-IDF), link graph, sessions and user paths |
| Data mining | Content mining (classification, clustering), structure mining (PageRank, HITS), usage mining (association rules, sequence patterns) |
| Evaluation | Measure interestingness, filter out trivial patterns (e.g. every user visits the home page) |
| Presentation | Visualise navigation paths, recommendations, reports |
Example: for an online shop, logs are cleaned and divided into sessions, association rules find that users who view "laptop" pages also view "laptop bag" pages, and the shop recommends bags on laptop pages. The process is iterative: results may lead to returning to earlier steps.
✓ Answer checked by Claude Opus 5.5 (high effort)
Related questions
- Web log mining and WWW mining architecture · Data Mining, chapter 7
- Web mining, importance and techniques · Data Mining, chapter 7
- Short note: WWW mining · Data Mining, chapter 7
- Data mining as a step in KDD · Data Mining, chapter 1
- Web structure/content/usage mining; privacy · Data Mining, chapter 7
- Web mining structure and challenges · Data Mining, chapter 7
- Web mining with structure mining · Information System, chapter 7
- Web mining importance and types · Information System, chapter 7 · asked 3×
- Text mining vs web usage mining · Information System, chapter 7
- Internet vs WWW; client-server paradigm · Web Technology, chapter 1