UNIT 5

1. What is Big Data?

➔
Big Data refers to large volumes of structured and unstructured data that cannot be
processed by traditional database systems.

2. Define Real-Time Data Mining.

➔
Real-Time Data Mining involves analyzing data as it arrives to provide immediate
insights or predictions.

3. What is Text Mining?

➔
Text Mining is the process of extracting useful information and patterns from
unstructured text data.

4. What is Web Mining?

➔
Web Mining is the use of data mining techniques to automatically discover and
extract information from web documents and services.

5. Define Stream Data Mining.

➔
Stream Data Mining involves processing and analyzing continuous, high-speed data
streams, such as sensor or financial transaction data.

6. What is Data Privacy in Data Mining?

➔
Data Privacy refers to protecting sensitive information during data mining, ensuring
individual data is not exposed or misused.

7. What is the purpose of Anomaly Detection in Data Mining?

➔
Anomaly Detection identifies unusual patterns or data points that do not conform to
expected behavior.
8. What is Sentiment Analysis?
➔
Sentiment Analysis determines the sentiment (positive, negative, neutral) expressed in
textual data, commonly used in social media mining.

9. What is the role of Machine Learning in Data Mining?

➔
Machine Learning provides algorithms that allow computers to learn from data and
make predictions without explicit programming.

10. What is the significance of Data Mining in Healthcare?

➔
Data Mining helps in disease prediction, patient diagnosis, and discovering hidden
patterns in medical records for better treatment decisions.

11. What is Clustering in the context of Big Data?

➔
Clustering groups similar data items from large-scale datasets to detect patterns, such
as customer segmentation in Big Data analytics.

12. Define Predictive Analytics.

➔
Predictive Analytics uses historical data and statistical algorithms to predict future
trends and behaviors.

13. What is the difference between Data Mining and Knowledge Discovery in
Databases (KDD)?
➔
Data Mining is a step within the KDD process, which includes data selection,
preprocessing, transformation, data mining, and interpretation of results.

1. Explain the concept of Big Data and its characteristics (3Vs).

➔
Big Data refers to massive volumes of data collected from various sources that
exceed the processing capability of traditional database systems.
The 3Vs of Big Data are:
o Volume: Large amount of data generated from social media, sensors,
transactions, etc.
 o Velocity: High speed at which data is generated and processed in real time
(e.g., streaming data).
o Variety: Different types of data (structured, semi-structured, unstructured)
like text, images, videos, logs.
These characteristics require specialized storage (e.g., Hadoop Distributed File
System) and processing frameworks (e.g., Apache Spark).

2. Describe Real-Time Data Mining and its applications.

➔
Real-Time Data Mining processes and analyzes data as it is generated, enabling
immediate insights and decision-making.
Applications:
o Fraud detection in financial transactions.
o Real-time recommendation systems (e.g., Netflix recommendations).
o Network intrusion detection.
Challenges include handling high velocity data and ensuring low-latency
processing.
Tools like Apache Flink and Apache Storm are often used for real-time data
mining.

3. Explain Text Mining and its steps with an example application.

➔
Text Mining extracts structured knowledge from unstructured text data through:
1. Text Preprocessing: Tokenization, removing stop words, stemming.
2. Text Representation: Converting into numerical format (e.g., TF-IDF vectors).
3. Pattern Discovery: Clustering, classification, association rules.
Example Application:
Sentiment analysis on customer reviews to classify them as positive or
negative.
Text mining is widely used in social media analytics, customer feedback
analysis, and document summarization.

4. What is Stream Data Mining and how does it differ from traditional Data
Mining?
➔
Stream Data Mining deals with continuous data flows (e.g., sensor data, financial
transactions), processing data incrementally.
Differences from Traditional Data Mining:

Stream Data Traditional Data

Feature
Mining Mining
Data Arrival Continuous Static Dataset
Limited (single
Memory Entire Dataset
pass)
 Stream Data Traditional Data
Feature
Mining Mining
Time Sensitivity Real-Time Batch Processing
Example: Real-time analysis of stock market
trends vs historical trend analysis.

5. Discuss Sentiment Analysis and its importance in Data Mining.

➔
Sentiment Analysis involves identifying subjective information in text, such as
opinions, emotions, or attitudes.
o Uses Natural Language Processing (NLP) and Machine Learning.
Importance:
o Businesses understand customer opinions.
o Monitors brand reputation in real time.
Example:
Analyzing Twitter data to determine public sentiment about a product launch.
Challenges include sarcasm detection, multiple languages, and ambiguous
expressions.

6. What are Anomaly Detection techniques and where are they applied?
➔
Anomaly Detection identifies data points that deviate from the norm.
Techniques:
o Statistical Methods: Z-score, Gaussian Distribution models.
o Distance-Based: K-NN-based outlier detection.
o Density-Based: DBSCAN.
Applications:
o Credit card fraud detection.
o Network intrusion detection.
o Fault detection in manufacturing.
Example: Monitoring network traffic to detect unusual spikes indicating an
attack.

7. Explain Data Privacy in Data Mining and techniques used to preserve it.
➔
Data Privacy ensures individual information is not exposed during mining.
Techniques:
o Data Anonymization: Removing personal identifiers.
o k-Anonymity: Ensures each data record is indistinguishable from at least k
others.
o Differential Privacy: Adds noise to data to prevent identification of
individuals.
Example: Health records can be anonymized before mining to prevent patient
identification while analyzing disease patterns.
8. Discuss the role of Machine Learning in advancing Data Mining research
frontiers.
➔
Machine Learning enables automated pattern discovery from data without explicit
programming.
o Algorithms such as Neural Networks, Decision Trees, and SVM enhance
predictive and classification tasks.
Impact on Data Mining:
o Improved handling of unstructured data (images, text).
o Scalability for big data.
o Adaptive models for streaming data.
Example: Deep Learning models like CNNs used for image-based medical
diagnosis.

9. Explain the importance of Data Mining in Healthcare with examples.

➔
Data Mining in Healthcare helps in:
o Predicting disease outbreaks.
o Diagnosing diseases from patient records.
o Drug discovery and patient monitoring.
Example:
Mining Electronic Health Records (EHR) to identify patients at risk of
diabetes based on historical data patterns.
This improves decision-making, reduces cost, and enhances patient care
quality.

10. Describe Predictive Analytics and its role in decision making.

➔
Predictive Analytics uses historical data and statistical models to predict future trends
and outcomes.
Steps:
1. Data Collection.
2. Data Cleaning and Preprocessing.
3. Model Building (e.g., Regression, Decision Trees).
4. Model Evaluation (accuracy, precision).
5. Deployment for decision-making.
Example:
Predicting customer churn in telecom to take proactive retention actions.
It helps businesses improve performance and reduce risks by anticipating
future behavior.
Q1. Explain the key applications of data mining in modern industries and society.
Answer:
Data mining is applied across multiple domains:

 Business: Market basket analysis, customer segmentation, sales forecasting, fraud

detection.
 Healthcare: Predicting diseases, patient diagnosis, personalized treatment plans.
 Education: Analyzing student performance, identifying at-risk students, curriculum
optimization.
 Finance: Credit scoring, risk management, detecting suspicious transactions.
 Social Media: Sentiment analysis, trend detection, user behavior prediction.

Impact on society: Data mining helps in informed decision-making, resource optimization,

social welfare programs, and discovering hidden patterns for public good. Ethical use ensures
privacy and fairness.

Q2. Describe data mining for sentiment and opinion analysis with examples.
Answer:

 Sentiment Analysis: Extracts subjective information from textual data (positive,

negative, neutral).
 Opinion Mining: Determines public opinion about products, services, or events.

Example:

 Amazon reviews: Classifying customer feedback as positive, negative, or neutral.

 Twitter sentiment analysis: Detecting public mood about elections or product
launches.

Techniques:

 Natural Language Processing (NLP), text mining, machine learning classifiers like
SVM, Naïve Bayes, and deep learning methods.
 Applications: Marketing strategies, political campaigns, brand monitoring.

Q3. Discuss data mining methodologies and systems.

Answer:
Methodologies:

 CRISP-DM: Business understanding → Data understanding → Data preparation →

Modeling → Evaluation → Deployment.
 KDD Process: Selection → Preprocessing → Transformation → Data mining →
Interpretation/Evaluation.

Data Mining Systems:

  Classification Systems: Predict categories (e.g., spam detection).
 Clustering Systems: Group similar objects (e.g., customer segmentation).
 Association Rule Mining: Discover relations among items (e.g., market basket
analysis).
 Hybrid Systems: Combine multiple techniques for better results.

Components: Database, data warehouse, data mining engine, pattern evaluation, user
interface.

Q4. Explain the role of structuring unstructured data for knowledge mining.
Answer:

 Unstructured Data: Text documents, social media posts, images, videos.

 Challenges: No predefined schema, complex formats, high dimensionality.
 Techniques for Structuring:
o Text mining and NLP for extracting entities, sentiments, and topics.
o Feature extraction (TF-IDF, word embeddings).
o Image/video mining using computer vision techniques.

Outcome: Structured representations allow efficient data mining and knowledge discovery.
For example, converting social media data into structured sentiment scores for analysis.

Q5. What is data augmentation? How does it enhance data mining?

Answer:

 Data Augmentation: Creating new training data from existing data to improve
machine learning models.
 Techniques:
o Text: Synonym replacement, paraphrasing, back translation.
o Images: Rotation, scaling, flipping, noise addition.
o Tabular Data: SMOTE for balancing classes.

Benefits in Data Mining:

 Increases dataset size and diversity.

 Reduces overfitting and improves model robustness.
 Enables better generalization in predictive analytics.

Q6. Explain the transition from correlation to causality in data mining.

Answer:

 Correlation: Measures statistical association between variables.

o Example: Ice cream sales correlate with drowning incidents.
 Causality: Establishes a cause-effect relationship.
 o Techniques: Randomized controlled trials, causal inference, Bayesian
networks, Granger causality.

Importance:

 Helps in making actionable decisions rather than misleading assumptions.

 Example: Marketing strategies based on causal effects rather than mere correlation.

Q7. Discuss the concept of networks as context in data mining.

Answer:

 Network Data: Social networks, citation networks, IoT networks.

 Importance: Relationships between entities provide valuable context.
 Techniques: Graph mining, link prediction, community detection, centrality
measures.

Applications:

 Social media analysis (influencer identification).

 Fraud detection in financial networks.
 Understanding protein-protein interactions in bioinformatics.

Q8. Describe privacy-preserving data mining techniques.

Answer:

 Goal: Protect sensitive information while still extracting useful knowledge.

 Techniques:
o Data anonymization (k-anonymity, l-diversity).
o Differential privacy (adding noise to datasets).
o Secure multi-party computation (collaborative mining without sharing raw
data).

Applications:

 Health data sharing, financial analysis, customer analytics while complying with
regulations (e.g., GDPR).

Q9. Explain human-algorithm interaction and mining beyond accuracy.

Answer:

 Human-Algorithm Interaction: Users guide algorithms with domain knowledge,

adjust parameters, and interpret outputs.
 Beyond Accuracy: Focus on fairness, interpretability, robustness.
o Fairness: Avoid bias in predictions (e.g., hiring decisions).
 o Interpretability: Transparent models for user trust.
o Robustness: Models resilient to noisy or adversarial data.

Significance: Ensures ethical, reliable, and socially responsible data mining.

Q10. How can data mining be used for social good? Provide examples.
Answer:
Data mining can improve societal outcomes:

 Healthcare: Predicting epidemics, optimizing treatment plans.

 Education: Identifying students at risk to provide early intervention.
 Environment: Climate pattern analysis, disaster prediction.
 Public Policy: Resource allocation, crime prediction, urban planning.

Example:

 Using mobile data to track disease outbreaks.

 Analyzing social media for disaster relief coordination.

Outcome: Data-driven decisions that benefit society, improve quality of life, and optimize
resource utilization.