05 - BI and Warehousing - slides

Course Information

• Institution: Mittuniversitetet, Mid Sweden University

• Course: Introduction to Business Intelligence

• Teacher: Ivika Jäger

• Date: November 2024

Managerial Decision Making

• Making decisions based on goals includes:

  • Historically: Intuition and experience used.

  • Modern Approach: Increasing focus on data-driven decisions.

• Importance of detailed on-demand reporting and advanced analytics to project potential outcomes.

Decision-Making Process

1. Intelligence: Define and analyze the problem.

2. Design: Develop models to test solutions.

3. Choice: Select the most viable solution.

4. Implementation: Execute the solution and monitor progress.

• Feedback and progress monitoring are crucial for success.

Effective Problem Solving

• Types of Problems:

  • Structured (well-defined)

  • Unstructured (complex)

• Strategies:

  • Decompose problems for easier management.

  • Assign ownership and authority appropriately.

Decision Support Systems (DSS)

• Definition: High-performance computer dashboard aiding in data analysis.

• Integrates data from various sources, enabling simulation of scenarios.

• Not limited to one specific function; encompasses business reporting, data analysis, etc.

Evolution of DSS

• 1970s: Focus on periodic reporting.

• 1980s: Growth of on-demand reporting due to digitization.

• 1990s: Introduction of BI systems and data warehousing.

• 2000s: Emergence of data mining and SaaS for accessible analytics.

• 2010s: Trends towards big data and real-time analytics.

• 2020s: Adoption of AI, deep learning, and IoT for automated analysis.

Business Intelligence (BI)

• Definition: Architecture, tools, and methodologies that provide insights through data manipulation and analysis.

• Focuses primarily on descriptive analytics to convert data into actionable information.

Data Analytics and Business Intelligence

• Types of Analytics:

  1. Descriptive: What happened?

  2. Predictive: What will happen?

  3. Prescriptive: What should I do?

• Utilization of dashboards, scorecards, and implementation of data mining for better strategic decisions.

Tools for Business Intelligence

• Popular BI tools include:

  • Excel: For data analysis.

  • SQL: For database management.

  • PowerBI: For visualizations.

  • Python and R: For data science applications.

Data Science vs. Analytics

• Analytics (BI): Reporting focused (Excel, SQL, PowerBI).

• Data Science: Predictions and optimization driven (Python, R).

• Note: Both fields complement each other.

Careers in Data

• Related roles include:

  • Analyst (business, strategic, quantitative)

  • Statistician/Researcher

  • BI/ML/AI Engineer

  • Data Scientist

  • Database Administrator

Data Warehousing

• Definition: Stores data for analysis and decision-making across multiple sources.

• Designed for large-scale analytical processing rather than transaction processing.

Benefits and Costs of Data Warehousing

• Benefits:

  • Flexible and faster decision-making.

  • Consolidation of corporate data.

  • Reduced processing costs.

• Costs:

  • Hardware, software, and network expenses.

  • Training costs for users.

Keys to Success in DW Implementation

• Prioritize user involvement in development.

• Ensure proper training and technical support.

• Maintain continuous feedback during the process.

Data Integration and Management

• ETL Pipeline:

  • Utilize APIs and web scraping for data extraction.

  • Clean and organize data before loading into the warehouse at scheduled intervals.

  • Maintain metadata and error management for system integrity.

Data Warehouse Administrator Role

• Management of both technical and business aspects of DW.

• Requires a mix of technical skills, business knowledge, and communication.

• AI may assist in documentation but cannot replace human oversight.

Data Lakes

• Definition: Store raw, unstructured data for flexible analysis.

• Ideal for technical users; complement data warehousing for non-technical users.