All subjects · Information Technology · WIT13 2 hour written exam on data, enabling technologies and systems development
Data mining, data warehousing and data analytics
Edexcel IAL Information Technology · WIT13 2 hour written exam on data, enabling technologies and systems development
Collecting Big Data is pointless unless it is turned into insight. This lesson covers data warehousing, data mining and the three types of analytics: descriptive, predictive and prescriptive (spec 12.3.4).
Data warehousing
A data warehouse is a large store of data brought together from many operational systems (sales, stock, finance, web) for analysis and reporting, not day-to-day transactions. Data is loaded through ETL: extract from source systems, transform (clean, remove duplicates, convert to common formats and codes), load into the warehouse. Warehouse data is integrated, historical (kept over years so trends can be seen) and mostly read-only. Smaller subject-specific subsets are called data marts.
Data mining
Data mining is the automated searching of large data sets to discover patterns, relationships and anomalies that were not previously known. Techniques include:
- Association (market basket analysis): items bought together. Measured by support (fraction of all transactions containing both items) and confidence (fraction of transactions with A that also contain B).
- Classification: putting records into known groups (likely to default or not).
- Clustering: finding natural groups without predefined labels (customer segments).
- Anomaly (outlier) detection: unusual records, such as fraudulent transactions.
- Regression and forecasting: predicting a numeric value from trends.
Three types of analytics
| Type | Question answered | Example |
|---|---|---|
| Descriptive | What happened? (and why) | Dashboards of last month's sales by region; reports of past fraud |
| Predictive | What is likely to happen? | Forecasting demand next month; predicting which customers will leave |
| Prescriptive | What should we do about it? | Recommending stock levels, prices or interest rate changes; route optimisation |
Each type builds on the one before; prescriptive is the most complex and valuable.
Applying it in an exam answer
The June 2024 mark scheme's indicative content for a Big Data question about a financial institution used exactly this structure: obtain the data (own records, bought or licensed data), store it (cloud storage, skilled staff to keep up with volume and velocity) and use it (descriptive analysis of past events such as fraud, predictive analysis of trends, prescriptive analysis to inform decisions such as changing interest rates, and pattern analysis to spot outliers).
Worked example
In 2000 supermarket transactions, 400 contain bread and 120 contain bread and butter. Calculate the support and confidence of the rule 'bread implies butter'.
- Support = 120 / 2000 = 0.06
- Confidence = 120 / 400 = 0.3
- So 6% of all baskets contain both, and 30% of bread buyers also buy butter
Answer: Support 0.06, confidence 0.3
Check yourself
- In 5000 transactions, 800 contain pasta and 200 contain pasta and sauce. What is the confidence of 'pasta implies sauce'?
Show answer
0.25. 200 / 800 = 0.25.
- Using the same figures (5000 transactions, 200 with pasta and sauce), what is the support of the rule?
Show answer
0.04. 200 / 5000 = 0.04.
- A bank's system recommends raising an interest rate by 0.25% to maximise profit while keeping customers. Which type of analytics is this?
- Prescriptive
- Descriptive
- Predictive
- Diagnostic only
Show answer
Prescriptive. It recommends an action.
- An online retailer estimates which customers are likely to cancel their subscription next month. Which type of analytics is this?
- Descriptive
- Prescriptive
- Data warehousing
- Predictive
Show answer
Predictive. It forecasts a future outcome.
- A dashboard shows the number of hospital admissions each week last year. Which type of analytics is this?
- Predictive
- Descriptive
- Prescriptive
- Clustering
Show answer
Descriptive. It summarises the past.
- During ETL, what happens at the transform stage?
- Data is copied out of the source systems
- Data is written into the warehouse tables
- Data is deleted from the source systems
- Data is cleaned and converted into consistent formats
Show answer
Data is cleaned and converted into consistent formats. Extract, transform (clean and convert), load.
- A company groups customers into segments with similar buying habits without deciding the groups in advance. Which data mining technique is this?
- Association
- Clustering
- Classification
- Regression
Show answer
Clustering. Groups found from the data itself.
- Explain the difference between data mining and descriptive reporting.
Show answer
Descriptive reporting summarises what is already known about what happened, whereas data mining automatically searches large data sets to discover new hidden patterns or relationships.. Mining discovers; reporting summarises.
Sources
- Pearson Edexcel International Advanced Level in Information Technology specification (first teaching 2018), Issue 2, February 2021
- Pearson Edexcel IAL Information Technology: scheme of work for Unit 3
- Pearson Edexcel IAL Information Technology WIT13/01 mark scheme, June 2024 series
Last updated 8 November 2026. Spotted a mistake? Tell us.