SUMMARY
Data Ontologies for Data-Driven Decision-Making: Development and Use

Transportation agencies are exploring tools and techniques to treat data as a business asset. This change arises from a shift toward performance-based management, which necessitates cross-cutting analytics and data-driven decision-making. Furthermore, technological advancements have led to agencies managing numerous legacy systems and a collection of disparate architectures. Although these legacy systems have limitations, they remain highly relevant because they are well understood by long-established experts who work closely with the data. Some agencies are taking decisive steps to enhance the functionality of legacy systems for cross-functional decisions by developing “data lake” or “data warehouse” approaches. However, advanced data representation and knowledge management techniques, such as developing a data ontology, could improve these strategies. A data ontology is a framework for characterizing and defining classes, attributes, and their relationships in a domain, providing a shared meaning across multiple users.

This guide equips transportation agencies with essential resources to develop and implement data ontologies that support data-driven decision-making. It serves stakeholders at all levels of a transportation agency, aiding them in understanding the value of data ontologies from both business and technical perspectives. This guide also emphasizes the organizational, functional, and technical needs for successfully integrating these techniques into practice.

This guide is based on a conceptual framework consisting of four cyclical strategies: (1) Designing a plan that enables success (D), (2) Assembling the ontology (A1), (3) Testing the ontology (T), and (4) Adapting to change (A2), including improving the ontology and its development process based on the lessons learned, changes in program objectives, or use case outcomes, collectively referred to as the DATA Framework. The strategies and supporting activities for these pillars include the following:

Pillar D: Designing a Plan

Step D.1 – Clarify the Purpose

Step D.2 – Confirm the Domain

Step D.3 – Classify Stakeholders

Step D.4 – Create a Plan

Pillar A1: Assembling the Ontology

Step A1.1 – Define the Scope

Step A1.2 – Gather Relevant Information

Step A1.3 – Organize and Classify the Information

Pillar T: Testing the Ontology

Step T.1 – Define and Execute Verification Cases

Step T.2 – Modify and Validate the Ontology

Step T.3 – Deploy the Ontology

Pillar A2: Adapting to Change

Step A2.1 – Create an Ontology Repository

Step A2.2 – Establish the Update Protocol

Step A2.3 – Identify Enhancements

Step A2.4 – Establish a Versioning Control Strategy

Step A2.5 – Implement Updates and Document Changes

Step A2.6 – Communicate the Updates

The following attributes describe the DATA Framework:

• Stronger emphasis on planning and governance—This attribute underscores the importance of initial planning in ensuring that the right people, sufficient funding, and all other necessary resources for success are identified.
• Actionable steps—The framework breaks down high-level strategies into specific, clear, and practical activities to achieve the userʼs goals. It includes a sample list of competency questions, key tips, and practical examples to ease execution.
• Simplicity—The frameworkʼs simplicity allows users with different experience levels and expertise to adopt and implement the methodology with minimal effort.
• Flexibility and scalability—The framework allows users to customize it to meet their needs and scale implementation without compromising its efficacy in addressing the ever-evolving data landscape.
• Continuous improvement—The frameworkʼs cyclical nature ensures a continuous feedback loop, incorporating lessons learned into each step to enhance the outcome.

This guide discusses typical barriers that can challenge its implementation and application. These implementation barriers include the following:

• Lack of leadership buy-in—A lack of leadership buy-in can hinder sustainable funding, weaken staff commitment, and undermine the authority to hold people accountable.
• Leadership changes and staff turnover—In an environment of continuous turnover coupled with the inevitable movement of staff from one office or position to another, this barrier must be appropriately managed to minimize potential setbacks.
• Resistance to change—Implementing this guide and framework will necessitate adjustments to existing practices and processes, which key stakeholders may resist. Developing strategies to manage this resistance can enhance success.
• Resource limitation—Developing data ontologies can be resource intensive, including sustainable funding, adequate support staff, availability of subject-matter experts (SMEs) to contribute meaningfully, and technological resources, such as tools and ontology editors. The lack of these resources can hinder implementation.

This guide provides direction to decision-makers and practitioners at transportation agencies on how to integrate the DATA Framework into their data governance practices to enhance knowledge management, improve data sharing, promote integration and semantic interoperability, and facilitate data-driven decision-making. It can be adopted at all levels of a transportation agency, including the strategic, program, and project levels. At the strategic level, agency leadership can foster institutionalization by establishing goals that encompass the adoption of data ontology principles. At the program level, mid-level managers can align operational activities, such as information technology modernization, asset management implementation, and transportation system management operations, with data ontology principles. At the project level, a data ontology can influence construction methods, such as digital delivery, by streamlining asset planning, design, and construction processes.

Lastly, the guide recommends adopting and maintaining change management strategies to facilitate the implementation of the DATA Framework. For example, some agencies may need to implement organizational, operational, or tactical changes to adapt to ontology development and application. Understanding the requirements to create an environment with the appropriate culture and necessary capacity can drive success.