世界经济论坛-移动数据运营的10条原则（英文）-2020.6-15页_870kb

10 Principles of Mobility Data Operationalization Summary

Core Content

This document outlines ten key principles for operationalizing mobility data, emphasizing the importance of thoughtful data collection, analysis, and application in the context of public and private mobility systems. It highlights the subjective nature of data, the need for comprehensive inventory management, the value of existing data frameworks, and the role of pilots and checks in ensuring robust and ethical data practices.

Main Principles and Key Points

Principle 1: Question your assumptions

• Data is subjective: Data reflects the biases and assumptions of those who collect and interpret it.
• Biases in data collection: For instance, tracking formal complaints for public transit safety overlooks unreported crimes and other safety concerns.
• Diversity in teams: Hiring individuals from diverse socioeconomic backgrounds helps challenge biases and improve data quality.
• Be ready to be wrong: Insights can challenge existing assumptions, and it's crucial to remain open to rethinking the project's approach.

Principle 2: Take inventory, and take it often

• Inventory is essential: You must understand what data you have and what you don’t.
• Dynamic data landscape: Data is constantly being collected and deleted, so continuous inventory is necessary.
• Standardization and integration: A framework for integrating and standardizing data is required, even for internal projects.
• User-friendly tools: For non-technical staff, tools should be accessible, easy to use, and legally compliant.

Principle 3: There is no need to reinvent the wheel

• Leverage existing frameworks: Many cities and organizations have already developed data-exchange standards.
• Avoid guesswork: Use established privacy regulations and best practices rather than creating your own.
• Research similar projects: Learn from past initiatives to avoid repeating mistakes and build on existing knowledge.

Principle 4: Data likes (selective) company

• Combine data sources: Different datasets can provide complementary information to answer complex questions.
• Identify relevant data: Not all data gaps are critical; some can be ignored without affecting the outcome.
• Cost considerations: Data collection and integration can be costly, but it's important to invest in quality data.

Principle 5: Data is nothing but insights are everything

• Focus on the question: Every data project should start with a clear, defined question and purpose.
• Cut through noise: Prioritize data that directly answers your question rather than being overwhelmed by irrelevant information.
• User-centric approach: Ensure all data and insights are relevant to the end user and their needs.

Principle 6: Less is most definitely not less

• Avoid overcomplication: Too much data can lead to confusion and inefficiency.
• Prioritize data pools: Stick to agreed-upon data sources to maintain focus and reduce complexity.
• Qualitative and quantitative balance: Both types of data are important and should be used in tandem.

Principle 7: Ambition has many traps

• Do not try to do everything: Ambitious projects require breaking down into smaller, manageable questions.
• Define parameters clearly: For example, accessibility can be defined by time rather than abstract concepts.
• Test one question at a time: Pilots and data analysis should be guided by specific, focused questions.

Principle 8: Pilots are your best friends and your worst enemies

• Pilots as data points: Pilots should be used to test insights, not as standalone projects.
• Avoid biased pilots: Pilots must reflect real-world conditions and not be influenced by external factors like funding or location.
• Adapt based on pilot results: Pilots can reveal unexpected insights or limitations, requiring flexibility and reevaluation.

Principle 9: Don’t forget your checks and balances

• Privacy is more than legal compliance: It's a cultural and personal issue that requires careful consideration.
• Include a critical voice: Have someone on your team who can question assumptions and identify flaws.
• Stay committed to your framework: Once data rules and privacy standards are set, avoid changing them without justification.
• Align with real-world needs: Ensure your project reflects the actual needs, realities, and governance gaps of the region.

Principle 10: Big questions are only as good as small realities

• Embrace ambiguity: Data projects often start with big questions but require grounding in real data.
• Adapt as you go: You may discover gaps or need new tools and software as you work with data.
• Impact over prowess: The value of data work lies in its real-world impact, not in technical complexity or speed.

Key Takeaways

• Data is inherently subjective and must be approached with awareness of biases and assumptions.
• Continuous inventory and standardization are crucial for effective data integration.
• Leverage existing frameworks and learn from past initiatives to avoid mistakes.
• Focus on user needs and prioritize meaningful insights over unnecessary data.
• Pilots should test insights, not be used as a means to demonstrate data collection.
• Checks and balances, including privacy considerations and critical feedback, are essential for robust data practices.
• Ambition must be tempered with patience and a step-by-step approach to avoid overwhelming complexity.
• Data work is a journey of discovery and adaptation, not a one-time task.