Business Intelligence Concepts

Business Intelligence (BI) Focus Areas

• Operational BI:

  • Objective: Manage daily operations; integrate BI with operational systems.
  • Primary Users: Managers, analysts, operational users.
  • Time Frame: Intraday.
  • Data: Real-time metrics.
  • Results: Immediate actions leading to sales revenue.

• Tactical BI:

  • Objective: Conduct short-term analysis to achieve strategic goals.
  • Primary Users: Executives, managers.
  • Time Frame: Days to weeks to months.
  • Data: Historical metrics.
  • Results: Daily analysis leading to refined campaign strategies.

• Strategic BI:

  • Objective: Achieve long-term organizational goals.
  • Primary Users: Executives, managers.
  • Time Frame: Months to years.
  • Data: Historical metrics.
  • Results: Planning leading to successful marketing campaigns.

Integration and Latencies

• Data Latency:

  • Definition: Time duration to make data ready for analysis; involves extracting, transforming, cleansing, and loading data into the database.

• Analysis Latency:

  • Definition: Time taken from data availability to completion of analysis.

• Decision Latency:

  • Definition: Time taken by humans to understand analytic results and decide on action.

Data Mining Overview

• Definition: Process of analyzing data to extract information not directly available from the raw data.

• Three Elements of Data Mining:

  • Data: Foundation for data-directed decision-making.
  • Discover: Identifying new patterns, trends, and insights.
  • Deployment: Implementing discoveries to drive organizational success.

• Data Mining Process Steps:

  1. Business Understanding:
  • Identify business goals, assess the situation, define goals, create a project plan.
  1. Data Understanding:
  • Analyze data for quality issues; gather, describe, explore and verify data.
  1. Data Preparation:
  • Select, cleanse, integrate, and format data for analysis.
  1. Data Modeling:
  • Apply mathematical techniques; select modeling technique, design tests, build models.
  1. Evaluation:
  • Analyze trends, assess business problem solution potential, evaluate results, determine next steps.
  1. Deployment:
  • Plan deployment, monitor implementation, analyze results, review final reports.

Data Concepts

• Data Profiling:

  • Collecting statistics about existing data sources.

• Data Replication:

  • Sharing information to ensure consistency among multiple data sources.

• Recommendation Engine:

  • A data mining algorithm that uses customer purchase actions to recommend complementary products.

Data Mining Techniques

• Estimation Analysis:

  • Determines values for unknown continuous variables or their estimated future values.

• Affinity Grouping Analysis:

  • Reveals relationships between variables, their nature, and frequency.

• Market Basket Analysis:

  • Uses purchasing data to identify consumer buying behavior and predict further purchases.

• Cluster Analysis:

  • Divides data into mutually exclusive groups, optimizing proximity within groups and distance between different groups.

• Classification Analysis:

  • Organizes data into categories for better effectiveness and efficiency.

Decision Support Systems (DSS) Techniques

• What-if Analysis:

  • Assesses the impact of changes in variables on models.

• Sensitivity Analysis:

  • Studies the impact on other variables when one variable changes repeatedly.

• Goal Seeking Analysis:

  • Identifies necessary inputs to achieve a predetermined goal.

• Optimization Analysis:

  • Seeks optimum values for a target variable through adjustments, considering constraints.

Predictive Analytics Models

• Prediction:

  • Statements about expected future occurrences (e.g., sales projections, employee turnover).

• Model Types for Predictions:

  • Optimization Model:
  • Seeks to make decisions as effective as possible by maximizing productivity or minimizing waste.
  • Forecasting Model:
  • Utilizes time series data for obtaining future forecasts.
  • Regression Model:
  • Estimates relationships among various variables; essential for business insights.

Agile Business Intelligence (Agile BI)

• Definition:

  • An approach integrating Agile software development with BI to enhance outcomes of BI initiatives.

• Benefits of Successful BI Implementation:

  • Single access point for information for all users.
  • BI across departments.
  • Availability of up-to-the-minute information.

• Benefits Breakdown:

  • Quantifiable Benefits:
  • Working time saved on reports and selling information.
  • Indirectly Quantifiable Benefits:
  • Improved customer service leading to increased business.
  • Unpredictable Benefits:
  • Results from user discoveries.
  • Intangible Benefits:
  • Enhanced communication, job satisfaction, and knowledge sharing.