Chapter 01 - Key points

1. Types of data organizations

• After defining the conceptual organization of the data, a physical organization must be chosen.
• This organization defines how the data will be structured and saved on disk.
• This organization is transparent to the user.
• We can define three organizations:
  • Structured Data (or databases),
  • Unstructured Data (or free-form files, text files),
  • Semi-Structured Data (which will be the subject of this module).
• Each organization meets specific needs; the right organization must be chosen for each use case.

2. Structured Data

• These are generally databases.
• A database is a large collection of structured information stored on a permanent medium.

2.1. Presentation

• Data has a well-defined structure.
• In the case of databases:
  • Data is organized into tables.
  • Each table has several columns.
  • The table contains records; these records all have the same structure (same number, types, and column order).
• The DBMS checks the data for conformity before inserting it.

2.2. Advantages of structured data

• Centralized data: no redundancy,
• Consistent data: application of constraints,
• Can handle large amounts of data,
• High-level operations.

2.3. Disadvantages of structured data

• Binary format specific to the DBMS,
• Difficult for humans to read in its native format,
• Difficult to read using another DBMS,
• Difficult to exchange,

3. Unstructured Data

• By unstructured data, we mean text files.
• This is the most basic type of data.

3.1. Presentation

• This is a string of characters without a clear structure.
• These files are edited using text editors like Notepad.
• Important: If a word processing program (like Microsoft Word) is used to edit a text file, you must explicitly specify its type as "Text File" when saving. Otherwise, Microsoft Word will use its own format (with the .docx extension), which is a binary format.

3.2. Advantages of unstructured data

• Simple,
• Easy to edit,
• Easy for humans to read,
• Easy to share.

3.3. Disadvantages of unstructured data

• Manipulation via basic level operations (like open and close).
• Difficult to use in automated processing: the lack of structure makes writing algorithms to process them very difficult.

4. Semi-Structured Data

• They aim to meet the new needs required in the web context:
  • The need for data exchange: which requires "open" data and not proprietary binary formats,
  • The need for a structure for the exchanged data: to be able to process and display it correctly to the user.