1. Predictive
2. Descriptive

1. Classification
2. Regression
3. Time Series Analysis
4. Prediction

1. Association Rules,
2. Clustering,
3. Summarization,
4. And Sequence Discovery

Targeted Marketing

Loan Payment Prediction

Healthcare Mangement

Mobile Service Providers

To improve its operations and customer experience, Amazon uses a variety of strategies in the context of data mining jobs. Recommendation systems, a subset of collaborative filtering in data mining, are one well-known use. Amazon uses user browsing habits, past purchases, and preferences to make tailored product recommendations. This data-driven strategy improves the advised items' accuracy, which raises client happiness and sales. By examining trends and correlations across large datasets, the recommendation system helps Amazon provide personalized recommendations on the homepage, in email alerts, and during the checkout process. By encouraging cross-selling and upselling opportunities, this data mining work not only improves the user experience but also plays a significant part in Amazon's broader business strategy, ultimately leading to the company success.

Finding patterns, trends, and insights in big databases is the process of data mining. Data mining techniques can be used in a variety of ways in the field of education to improve administrative, instructional, and learning processes. Data mining, for instance, can support adaptive learning systems, which modify course materials and activities to suit the requirements and preferences of specific students. Through the analysis of student interactions with online learning resources like simulations, assignments, and quizzes, these systems are able to modify the content delivery, pacing, and difficulty level dynamically in order to maximize learning results.

Now, let's discuss some of the additional examples of applications for various data mining tasks:

Classification:

Example: Imagine you have a big pile of fruits, some are apples and some are oranges, but they all look similar. Using classification, we can teach a computer to tell the difference between apples and oranges by looking at their color, shape, and size. This way, it can sort them correctly, like a smart sorting machine at a fruit factory.

Clustering:

Example: Think about organizing your toys into different boxes based on how they look or what they do. Clustering is like that it groups similar things together. For instance, it can put all your toy cars in one box, dolls in another, and puzzles in another, making it easier to find them when you want to play.

Association Rule Mining:

Example: Suppose you always eat chips when you watch movies. Association rule mining helps us find patterns like this. It can discover that people who buy popcorn also tend to buy soda, so a movie theater might put them together in a combo deal, making customers happy and selling more snacks.

Regression:

Example: Let's say you want to know how fast a car can go based on its weight and engine power. Regression helps us predict things like that. By looking at data from different cars, it can tell us how much speed we can expect from a car with specific weight and engine power, just like a super-smart speed calculator!

Sequential Pattern Mining:

Example: Picture a day in your life from waking up to going to bed. Sequential pattern mining helps us understand the order of things we do. It might notice that you always brush your teeth before going to bed or have breakfast after getting dressed. This helps us understand routines and habits.

As a data mining consultant for an internet search engine company, I can gaurantee data mining techniques can be very very useful in various aspects of the company's operations. SO, some examples of how techniques such as clustering, classification, association rule mining, and anomaly detection can be applied are as follow:-

Clustering - By applying clustering algorithms to search query data, you can group together similar queries made by users. It easily identifies patterns and trends according to the customer behavior and, allow the search engine to provide more accurate and personalized search results.

Classification - Using classification algorithms can help the search engine to predict the intent behind user queries, such as informational, transactional, or navigational. This information can then be used to target advertisements more effectively. For example, Suppose if a user's query is classified as transactional (e.g., "buy laptop"), the search engine then can display relevant ads from advertisers selling laptops.

Association Rule Mining - By employing association rule mining techniques we can identify on user search and click-through data, the search engine can identify the customer behavior, such as frequent sequences of queries or items clicked. These patterns can be used to build recommender systems that suggest related queries or items to users based on their search history. For example, if a user A search for "smartphone" often and, he looks for "phone cases," then search engine can recommend phone cases to users B when he search phone.

Anomaly Detection – By this technique we can identify unusual or suspicious behavior on the search engine platform, such as fraud on click or other spam content. If there is a sudden spike detected in a click on a specific advertisement, then this technique could indicate click fraud, prompting the search engine to take action to mitigate the issue and protect advertisers' interests.

By these data mining techniques, the internet search engine company can enhance various aspects of its platform and improve overall performance and user satisfaction related to the company.

Sandeep Adhikari

CIHE22230

CIHE-23459

SUSHIL GHIMIRE

a> This is not a data mining task. A simple database query is required to complete this task.

b> This is not a data mining task. A simple database query would provide the desired information and then a finance or accounting task is required to determine the groupings to use to divide the customers. However, predicting the profitability of a new customer would be data mining.

c> This is not a data mining task. This is a accounting task to sum the sales of the company for financial reporting. It is a simple accounting.

d> This is not a data mining task. A simple database query could be used to collect this information and sort by student ID numbers.

e> This is not a data mining task since it is mathematically proven that we could use statistical methods to show there is a 1 out of 6 chance of rolling one of the numbers on the die.

f> This is a data mining task. A model could be built that is capable of predicting the continuous value of the stock price. includes the various factors that influence the price of a stock. Predicting time-series could definitely assist in the prediction of expected future stock price.

g> This is a data mining task. Data could be collected on normal and abnormal heart conditions to determine the heart rate of a patient for abnormalities. A model of the normal behavior of heart rate can be built that raises an alarm when an unusual heart behavior occurred. This would involve the area of data mining known as anomaly detection.

h> This is a data mining task. Similar to the answer in g, data could be gathered and then we can work on a model that is capable of labeling the activity as normal or abnormal.

i> This is not a data mining task. This is a simple signal processing.

Internet search engine companies have pioneered the use of data mining to provide a "free" service in exchange for users supplying behavioral data such as the information that they type in a search engine. This data can be utilized in a variety of ways to understand the Fields 5 behaviors of users which can be utilized for revenue generating activities such as advertising, upselling and cross-selling.

a> Yes, the census information contains information on race, ethnicity, medical conditions, and financial information which some people might be uncomfortable sharing.

b> Yes, it is because of the user information being released that can be utilized by hackers to impersonate others or black mail users.

c> Neutral, because as of now this does not present a concern today with the current technology. However, if more detailed data is available in the future this could be a cause for concern.

d> No, since the publicly available names and addresses in the telephone book don't present data privacy issues.

e> No, same as d.

Mahad Ahmed

CIHE231003

Customer segmentation: clustering into several groups based on purchasing behavior.

Stock Price Prediction:(Regression)predicts according to stock prices based on historical data and stocks.

And soo...on

Examples of Data Mining

1. Online Shopping Recommendations – Amazon and Netflix suggest products or movies based on your past searches and purchases.

2. Fraud Detection – Banks use data mining to detect unusual transactions and prevent fraud.

3. Healthcare Predictions – Hospitals analyze patient records to predict diseases and recommend treatments.

4. Social Media Trends – Facebook, Instagram, and Twitter analyze user activity to show relevant ads and trending topics.

5. Customer Support Chatbots – Companies use data mining to train AI chatbots to answer customer questions based on previous conversations.

For eg- A set of data collected in a restaurant shows that number of walk-in guests in public holidays are higher than on normal working days - that way more people can be rostered on public holidays to keep up with the crowd for sales

-Me

Key concepts of data mining are:

• Pattern recognitions

• Exploratory Data analysis (EDA)

• Data processing

Types of data mining:

• Descriptive data mining- finding patterns and describing data that summarize its characteristics.

• Predictive data mining- to predict future outcomes based on the data we have and its history.

Bishal Siris (CIHE22619)

The data mining process is breakdown into four steps:

• Data is collected and loaded in a data warehouse on-site or in a cloud service.

• Business analysts and management teams access data and determine how to organize that data.

• Custom applications software sorts and organizes the data.

• The end-users present data in an easy-to-share format, such as a graph or table.

-The most popular data mining techniques include association rules, classification, clustering, decision tree, K-nearest Neighbour, neural networks, and predictive analysis.

Data mining process

• Understand the business

• Understand the data

• Prepare the data

• Build the model

• Evaluate the results

• Implement changes and monitor

-Different data mining processing models have different steps but the general process is almost the same.

Some applications for data mining are:

• Sales

• Marketing

• Manufacturing

• Fraud detection

• Human resources

• Customer service

Pros of data mining

-It gives profitability and efficiency.

-Can be applied to any type of data and business problem

-It can reveal hidden information and trends

Cons of data mining

-It is complex

-Results and benefits are not guaranteed

-It can be expensive

Even social media companies use data mining to commodify their users to generate profits. Platforms like Facebook, TikTok, Instagram, and X, etc.

Some examples are:

• e-Bay and e-commerce

• Facebook-Cambridge Analytica Scandal

Rahul K C (CIHE22533)

Some real-world applications are as follows:

1. Classification

Definition: Assigns items in a dataset to predefined categories or classes.

Applications & Examples:

• Banking & Finance: Predicting whether a customer will default on a loan.

• Healthcare: Diagnosing diseases from patient data (e.g., diabetes prediction).

• Marketing: Classifying customers as likely or unlikely to respond to a campaign.

2. Regression

Definition: Predicts a numeric value based on input variables.

Applications & Examples:

• Real Estate: Predicting house prices based on features like location, size, and age.

• Weather Forecasting: Predicting temperature or rainfall.

• Sales Forecasting: Predicting future sales based on historical trends.

3. Clustering

Definition: Groups similar data items together based on similarity measures without predefined labels.

Applications & Examples:

• Market Segmentation: Grouping customers by purchasing behavior.

• Document Clustering: Organizing news articles or research papers.

• Image Segmentation: Identifying regions in medical images.

4. Association Rule Mining

Definition: Finds interesting relationships (associations) between items in large datasets.

Applications & Examples:

• Retail: Market basket analysis (e.g., people who buy bread often buy butter).

• Web Usage Mining: Recommending websites based on browsing patterns.

• Healthcare: Identifying frequently co-occurring symptoms and diseases.

5. Anomaly/Outlier Detection

Definition: Identifies unusual patterns that do not conform to expected behavior.

Applications & Examples:

• Fraud Detection: Detecting unusual credit card transactions.

• Network Security: Detecting cyberattacks or intrusions.

• Manufacturing: Identifying defective products.

6. Sequential/Temporal Pattern Mining

Definition: Finds patterns where one event follows another over time.

Applications & Examples:

• Retail: Identifying customer purchase sequences.

• Healthcare: Tracking progression of disease stages.

• Telecommunications: Predicting call drop sequences or network failures.

7. Text Mining / Sentiment Analysis

Definition: Extracts patterns or insights from textual data.

Applications & Examples:

• Social Media Analysis: Determining public sentiment about a product or politician.

• Customer Feedback Analysis: Extracting insights from reviews or surveys.

• Legal Document Analysis: Identifying key clauses or trends.

• If it contains words like “free” or “win”, the system classifies it as Spam.

• Otherwise, it classifies it as Not Spam.

Real-World Uses

• Email spam filtering

• Disease diagnosis in healthcare

• Credit approval in banks

• Fraud detection

Simple idea:
Classification predicts which category an item belongs to based on past data.

CIHE240909

Main uses:

• User behavior analysis – understanding what content people like.

• Targeted advertising – showing ads based on user interests by Meta Platforms.

• Trend detection – identifying popular hashtags and topics.

• Influencer analysis – measuring engagement and follower growth.

• Spam detection – identifying fake accounts and bots.

Instagram data mining helps businesses and researchers understand user activity and improve marketing strategies.

Cihe240892

Aashish Karki

Data Mining Task: Classification

The example for this task is an email services automatically filtering spam emails. The system analyzes email content and classifies messages as spam or non-spam.

Diksha Khatri

CIHE240714

1. Classification

Classification is a data mining technique used to assign data into predefined categories based on patterns learned from historical data.

Examples:

• Email spam detection: Emails are classified as spam or not spam.

• Medical diagnosis: Patients are classified as healthy or having a disease based on medical test data.

• Loan approval systems: Banks classify applicants as low-risk or high-risk borrowers.

  
  

2. Clustering

   
   

   Clustering groups similar data objects together without predefined labels.

   
   

   Examples:

   • Customer segmentation: Businesses group customers based on purchasing behaviour to target marketing campaigns.

   • Social media analysis: Users are grouped according to their interests or activities.

   • Image analysis: Similar images or pixels are grouped together for pattern recognition.

     
     

3. Regression

   
   

   Regression is used to predict continuous numerical values based on historical data.

   
   

   Examples:

   • House price prediction: Predicting the price of houses based on factors such as location, size, and number of rooms.

   • Sales forecasting: Predicting future product sales using past sales data.

   • Weather prediction: Estimating temperature or rainfall levels.

4. Association Analysis

   
   

   Association analysis identifies relationships between items that frequently occur together.

   
   

   Examples:

   • Market basket analysis: Customers who buy bread often also buy butter.

   • E-commerce recommendations: Online stores suggest products that are frequently purchased together.

   • Website analysis: Identifying webpages that are often visited together.

   CIHE 240612

   Nirvik Niraula

• Classification: For example it is used to sort data into categories, like identifying emails as spam or not spam.

• Regression is used to predict numbers, such as predicting house prices or future sales.

• Clustering is used to group similar data together, like grouping customers with similar buying habits.

• Association is used to find relationships between items, such as customers buying bread often also buying butter.

• Outlier Detection is used to find unusual or abnormal data, like detecting fraud in bank transactions.

1. Real world example:

A supermarket uses data mining to analyze customer purchase data. It may discover that customers who buy bread often buy butter and milk at the same time. The store can then place these products close together or offer bundle discounts to increase sales.

The predictive power of data mining has changed the way business strategies are designed. It is now possible to understand the present in order to anticipate the future. Here are some examples of data mining in today's industry:

Agriculture

Agricultural companies can use data mining or data analysis to optimize crop conditions, helping to improve productivity and crop quality. In this case, crops undergo constant analysis of climatic variations and soil, water, and topography conditions, with the aim of predicting crop productivity and growth and detecting plant diseases, among other purposes.

Data Mining Task: Prediction e Classification:

Marketing

Data mining is used to explore ever-larger databases and improve market segmentation. By analyzing the relationships between parameters such as customer age, gender, tastes, etc., it is possible to predict their behavior in order to target personalized loyalty or acquisition campaigns. Data mining in marketing also predicts which users are likely to cancel a service, what interests them based on their searches, or what should be included in an email list to achieve a higher response rate.

Data Mining Task: Association Rule Mining

Retail

Supermarkets, for example, use joint purchase patterns to identify product associations and decide how to place them in different aisles and on different shelves. Data mining also detects the offers that are most valued by customers or that increase sales at the checkout line.

Data Mining Task: Anomaly Detection e Classification

Banking

Banks use data mining to better understand market risks. It is commonly applied to credit ratings and intelligent anti-fraud systems to analyze transactions, card movements, purchasing patterns, and customer financial data. Data mining also allows banks to learn more about our preferences or habits on the Internet to optimize the return on their marketing campaigns, study the performance of sales channels, or manage regulatory compliance obligations.

Data Mining Task: Prediction e Classification

Medicine

Data mining enables more accurate diagnoses. By having access to all patient information—medical history, physical examination, and previous treatment patterns—it is possible to prescribe more effective treatments. It also enables more effective, efficient, and economical management of healthcare resources by identifying risks, predicting diseases in certain segments of the population, or forecasting the length of hospital stays. Detecting fraud and irregularities and strengthening ties with patients by deepening knowledge of their needs are also advantages of using data mining in medicine.

Disease prediction – identifying the risk of diseases early, such as Diabetes.

Diagnosis support – helping doctors make accurate diagnoses using patient data.

Treatment improvement – analyzing which treatments work best for patients.

Patient monitoring – tracking patient health and detecting problems early.

Hospital management – improving hospital services and resource planning.

Healthcare data mining helps doctors, hospitals, and researchers improve patient care and make better medical decisions. CIHE23253

"ANOMALY DETECTION"

• is the technique of identifying rare events or observations which can raise suspicions by being statistically different from the rest of the observations.

• Such “anomalous” behavior typically translates to some kind of a problem like credit card fraud, a failing machine, or a cyber attack. In finance, with thousands or millions of transactions to watch, anomaly detection can help point out where an error is occurring, enhancing root cause analysis and quickly getting support on the issue.

• Anomaly detection helps the monitoring cause of chaos engineering by detecting outliers and informing the responsible parties to act.

• Machine Learning and AI are increasingly being used for anomaly detection for fraud detection and Anti-Money Laundering (AML).

Source: What is Anomaly Detection? ML Methods | Databricks. (n.d.). Databricks. https://www.databricks.com/blog/what-is-anomaly-detection

CIHE23427

1) The Data : Scientists use LiDAR( Light Detection and Ranging). They fly plane over a jungle and fire billions of laser pulses at ground which creates massive 'point cloud' of data.

2) The 'Mining' part : The computer uses Pattern Recognition to shift through the billions of data points. It 'mines' for specific shapes- straight line, perfect rectangles, and a flat squares which dont usually occur in nature.

3) The Discovery : In 2018, researchers mined LiDAR data from Guatemalan jungle. They expected to find a new ruins, but the data mining revealed a 'Megacity' of over 60,000 hidden Mayan structures that were invisible to people standing right on top of them.

CIHE231409