Module 1

1. Compare structured and unstructured data

Aspect	Structured Data	Unstructured Data
Format	Pre-defined schema (rows, columns), relational database tables	No fixed format, cannot be organized in tables
Storage & Management	Easier to store, manage with legacy solutions, requires less storage	Challenging to store/manage, requires more storage
Data Type Examples	Numbers, dates, strings, spreadsheets, CRM, sales, finance	Images, audio, video, word documents, emails, social media, surveys, multimedia
Share of Enterprise Data	About 20% (Gartner estimate)	About 80% (Gartner estimate)
Processing	Can use SQL and conventional analytics; query performance usually high	Cannot use standard SQL, requires advanced tools for extraction and analysis
Flexibility	Rigid structure: less flexible, more consistent	Highly flexible, adaptable to any content or format
Business Use	Transactional data, operational reports	Sentiment analysis, customer interactions, multimedia archiving
Querying	Easily queried	Difficult to query, need special processing

Structured data is vital for business transactions and quick analytics, while unstructured data holds rich contextual information but is complex to extract, analyze, and leverage without specialized technologies and approaches (e.g., machine learning, NLP, computer vision).

Reference: Pages 2–6

2. Characteristics of unstructured data

Unstructured data displays several distinct, critical characteristics:

No Predefined Structure: It does not conform to fixed formats or schemas, accommodating free text, images, videos, and other data types.
Extreme Flexibility: Can store a wide variety of content (text, multimedia, sensor data) without constraints, making it very adaptable to changes [page 7].
Heterogeneity: Produced in many formats and sources—websites, emails, social media, sensors (multilingual, multimodal, and multi-source) [page 6].
Large Volume: Generated at massive scale, overwhelming legacy storage and management systems [page 6].