JNTU B.Tech III DATA SCIENCE Previous Question Paper (February - 2022 - R18)

Data is primarily categorized based on its structure, nature, and the way it is processed. The three main types are structured, semi-structured, and unstructured data.

1. Structured Data
Structured data is highly organized and formatted into a fixed schema, typically residing in relational databases (RDBMS) where it can be easily accessed and queried using SQL. It follows a tabular format with rows and columns.

Example: A customer database containing fields like CustomerID, Name, DateOfBirth, and AccountBalance.

2. Semi-Structured Data
Semi-structured data does not reside in a relational database but contains organizational properties that make it easier to analyze. It uses tags, markers, or metadata to separate semantic elements and enforce hierarchies of records and fields.

Example: JSON (JavaScript Object Notation) or XML files used for web APIs, where data is stored in a nested key-value pair format.

3. Unstructured Data
Unstructured data lacks a specific pre-defined data model or organization. It is complex to process and requires advanced techniques like Natural Language Processing (NLP) or Computer Vision to extract insights. It accounts for the vast majority of data generated today.

Example: Social media posts, images, audio files, video recordings, and PDF documents.

Additionally, data can be classified by its Measurement Scale:

• Nominal: Qualitative data used for labeling variables without quantitative value (e.g., Eye color: Blue, Brown, Green).
• Ordinal: Data with a natural order or rank, but the intervals between values are not known (e.g., Customer satisfaction ratings: Poor, Neutral, Excellent).
• Interval: Numerical data where the difference between two values is meaningful, but there is no true zero point (e.g., Temperature in Celsius).
• Ratio: Numerical data with a clear definition of zero, allowing for meaningful ratios (e.g., Weight, Height, or Income).

Data collection is a systematic process of gathering information from various sources to analyze, evaluate, and derive insights. The process involves the following key steps:

1. Define the Objective: Clearly articulate the research question or the problem statement to determine what specific data is required.
2. Identify Data Sources: Determine whether the data will be primary (collected firsthand through surveys, interviews, or experiments) or secondary (retrieved from existing databases, journals, or reports).
3. Select Data Collection Methods: Choose appropriate techniques such as web scraping, API integration, surveys, or sensor data logging based on the nature of the data.
4. Design Data Collection Instruments: Develop the necessary tools, such as questionnaires, scripts, or data extraction templates, ensuring they are structured to minimize bias.
5. Data Collection Execution: Implement the collection process while ensuring data integrity, consistency, and adherence to ethical standards.
6. Data Validation and Cleaning: Verify the collected data for accuracy, completeness, and relevance, removing duplicates or erroneous entries before storage.

Various tools facilitate this process depending on the source and format of the data:

• Web Scraping Tools (e.g., BeautifulSoup, Scrapy, Selenium): These are used to extract unstructured data from websites. BeautifulSoup is ideal for parsing HTML, while Selenium is used for automating interactions with dynamic, JavaScript-heavy pages.
• API Clients (e.g., Postman, Requests library): APIs provide structured access to data from platforms like Twitter or Google Maps. The Python 'requests' library is the standard tool for making HTTP calls to these endpoints.
• Survey and Form Tools (e.g., Google Forms, SurveyMonkey): These platforms are essential for collecting primary qualitative and quantitative data directly from users through structured questionnaires.
• Database Management Systems (e.g., MySQL, MongoDB): SQL databases are used to collect and store structured data, while NoSQL databases like MongoDB are effective for gathering semi-structured or unstructured data from diverse sources.
• IoT and Sensor Data Tools (e.g., Apache Kafka, MQTT): For real-time data collection from physical devices, message brokers like Kafka and protocols like MQTT are used to stream and ingest high-velocity data into a central repository.
• Open Data Platforms (e.g., Kaggle, UCI Machine Learning Repository): These serve as secondary data collection tools, providing pre-cleaned datasets for research and modeling purposes.