Part A

Module 1 : Data Mining

2020 March
1. What do you mean by a transactional database?
Transactional database in data mining refers to a database system that records individual
transactions, such as purchases or reservations, and is commonly used to analyze patterns
and relationships in large volumes of data.

2. What is a concept hierarchy? Give an example.

Concept hierarchy in data mining refers to a hierarchical organization of related concepts
or categories, with each level of the hierarchy representing a different level of abstraction or
generalization. For example, in a hierarchy of animal species, the top level may be
"Animals," the second level may be "Mammals," the third level may be "Carnivores" and so
on, with each subsequent level representing a more specific category.

3. What is background knowledge? Give an example.

Background knowledge refers to information that is known about the data or domain being
analyzed and can be used to inform the mining process or interpret the results. For example,
in analyzing customer purchasing patterns, background knowledge about seasonal trends or
marketing campaigns could be used to help identify relevant patterns in the data.

2021 April
1. What do you mean by data mining?
Data mining is the process of discovering patterns, relationships, and insights from large
volumes of data, using statistical and machine learning techniques to identify hidden
patterns or knowledge.

2. What do you mean by interestingness?

Interestingness refers to the degree to which a discovered pattern or relationship is novel,
valid, useful, and understandable to the domain expert.

3. List two methods for dimensionality reduction.

● Principal Component Analysis (PCA): A statistical method that identifies the most
important variables in a dataset and reduces the number of dimensions by projecting
the data onto a new coordinate system based on the principal components.
 ● t-SNE (t-Distributed Stochastic Neighbor Embedding): A nonlinear dimensionality
reduction technique that is particularly useful for visualizing high-dimensional
datasets by preserving the local structure of the data while also revealing global
patterns and relationships.

2022 April
1. What is a multimedia database?
Multimedia database is a database that stores multimedia data such as images, audio, and
video, and allows efficient retrieval of this data.

2. Name different methods by which a classification model can

be represented.
A classification model can be represented using various methods such as decision trees,
rule-based systems, neural networks, support vector machines (SVM), and k-nearest
neighbor (k-NN) algorithms.

3. What is numerosity reduction?

Numerosity reduction is a process of reducing the number of data instances or objects in a
dataset while preserving the important characteristics and relationships between the data
points, often used for reducing the computational complexity of data mining algorithms.
Part B
Module 1 : Data Mining

2020 March
13. Explain data discretization and concept hierarchy
generation.
Data discretization is the process of converting continuous numerical data into categorical
data by partitioning the range of values into intervals, or bins, and assigning each value to
the corresponding interval. This technique is used to simplify data analysis and reduce the
number of variables in a dataset.

Concept hierarchy generation, on the other hand, is the process of organizing categorical
data into a hierarchical structure of concepts or categories based on their relationships, such
as generalization or specialization. This technique is used to create a meaningful and
organized representation of categorical data for analysis and decision-making.

2021 April
13. Differentiate classification and prediction.
Classification and prediction are two fundamental tasks in data mining that involve building
models to predict the class or value of a target variable based on a set of input variables.
The main difference between classification and prediction is the type of target variable.

In classification, the target variable is a categorical variable, and the goal is to predict the
class or category of the target variable based on the input variables. Examples of
classification include predicting whether a customer will churn or not, or whether a tumor is
malignant or benign.

In prediction, the target variable is a continuous numerical variable, and the goal is to
predict the value of the target variable based on the input variables. Examples of prediction
include predicting the price of a house or the revenue of a business.

2022 April
13. Explain the concept of data integration.
Data integration is the process of combining data from multiple sources into a single,
unified view that can be used for analysis and decision-making. This process involves
identifying and resolving any inconsistencies or conflicts in the data, such as differences in
data formats, units of measurement, or data structures, to ensure that the data is accurate
and complete.

Data integration is a crucial step in data mining, as it enables analysts to work with a larger
and more diverse set of data, and to gain insights that may not be possible with individual
data sources. Some common techniques for data integration include data warehousing,
which involves storing and organizing data from multiple sources in a centralized repository,
and data fusion, which involves combining data from multiple sources to create a more
comprehensive and accurate representation of the underlying phenomenon.
Part C
Module 1 : Data Mining

2020 March
22. Explain why the data needs to be preprocessed before
mining.
Data preprocessing is a crucial step in the data mining process. It involves transforming
raw data into a clean and structured format that can be analyzed to extract meaningful
insights. There are several reasons why data preprocessing is necessary before data
mining:

● Data quality improvement: Raw data may contain errors, inconsistencies, missing
values, outliers, and noise that can affect the accuracy of the analysis. Data
preprocessing helps to identify and correct these issues, resulting in improved data
quality.

● Data integration: Data may be stored in different formats, sources, and structures.
Data preprocessing helps to integrate data from different sources into a common
format, making it easier to analyze.

● Data reduction: Raw data may contain a large number of attributes, some of which
may be irrelevant or redundant for analysis. Data preprocessing helps to reduce the
dimensionality of data by selecting relevant attributes, resulting in faster and more
accurate analysis.

● Data normalization: Raw data may be expressed in different units and scales. Data
preprocessing helps to normalize data by scaling it to a common range, making it
easier to compare and analyze.

● Data transformation: Raw data may not be suitable for analysis using certain
algorithms or models. Data preprocessing helps to transform data into a suitable
format for analysis.

Overall, data preprocessing is essential for accurate and efficient data mining. It helps to
improve data quality, reduce noise, integrate data from different sources, reduce
dimensionality, normalize data, and transform data into a suitable format for analysis.
2021 April
22. Explain major issues in data mining.
Data mining, despite its immense potential, is a complex process fraught with challenges
and issues. Below are some of the major issues in data mining:

● Data Quality: The quality of the data being analyzed is a critical factor in the success
of any data mining project. The data must be clean, consistent, and accurate to
ensure that the results are meaningful and actionable. However, data from various
sources may contain missing or incorrect values, outliers, or noise, which can impact
the accuracy and validity of the results.

● Scalability: With the explosion of data in recent years, the volume of data to be
processed by data mining algorithms has increased exponentially. This increase in
data volume can be challenging for algorithms that are not designed to handle such
large data sets. Therefore, scalability is a major issue in data mining that needs to be
addressed to ensure that the algorithms are efficient and can handle large data sets.

● Data Privacy and Security: Data mining involves the use of sensitive data, such as
financial records or medical records, which may be subject to privacy laws or
regulations. Therefore, data privacy and security are crucial issues that must be
addressed to ensure that the data is not misused or compromised.

● Interpretability: Another major issue in data mining is the interpretability of the

results. Data mining algorithms often generate complex models that may be difficult
to interpret or understand by non-experts. Therefore, it is important to ensure that the
results of data mining are presented in a way that is understandable and actionable
by decision-makers.

● Algorithmic Bias: Data mining algorithms may be subject to algorithmic bias, which
is the tendency of algorithms to favor certain groups or individuals over others.
Algorithmic bias can result in unfair or discriminatory outcomes, which can have
serious consequences for the individuals or groups affected.

● Ethics: Data mining involves the collection and use of data, which can raise ethical
concerns. For example, the use of data mining for surveillance or profiling may be
considered unethical or illegal in some contexts. Therefore, ethical considerations
must be taken into account when designing and implementing data mining projects.

In conclusion, data mining is a powerful tool for uncovering insights and patterns in large
data sets, but it also poses several challenges and issues. Addressing these challenges is
crucial to ensure that the results of data mining are accurate, reliable, and actionable.
2022 April
22. Explain various data mining task primitives.
Data mining task primitives are the basic building blocks of the data mining process, which
define the type of patterns that can be mined from a dataset. There are several data mining
task primitives that are widely used in the field of data mining. Some of the important task
primitives are:

1. The set of task-relevant data to be mined: This refers to the portion of the
database that the user is interested in. It could include specific attributes, dimensions
of interest in a data warehouse, or any other relevant data that the user wants to
extract insights from.

2. The kind of knowledge to be mined: This refers to the specific function or analysis
that the user wants to perform. For example, the user may want to perform
classification, clustering, or association analysis on the data.

3. The background knowledge to be used in the discovery process: This refers to

any prior knowledge that the user has about the data, which can be used to improve
the accuracy and relevance of the data mining results. For example, the user may
have information about certain relationships or dependencies in the data, which can
be used to guide the mining process.

4. The interestingness measures and thresholds for pattern evaluation: This refers
to the criteria that are used to determine the usefulness or significance of the
patterns discovered during the data mining process. For example, the user may set a
threshold for the minimum support level or confidence level of association rules that
are considered interesting.

5. The expected representation for visualizing the discovered patterns: This refers
to the form in which the user wants to visualize the patterns that are discovered. This
could include various forms such as tables, graphs, charts, decision trees, or cubes.
The visualization is an important aspect of the data mining process as it can help the
user to better understand and interpret the results.

By using these data mining task primitives, different types of patterns can be identified in the
data. These patterns can help in making informed decisions and improving the overall
efficiency of the process.