Kasetsart University Research Intelligence

KU Urban Decision Intelligence

A system-design perspective for connecting data governance, human wisdom, and AI-assisted strategic intelligence at Kasetsart University

Back to home

About

Why this system was designed

This system is not intended to be only a researcher profile page. It is designed as a prototype platform that helps the university see the potential of people, work, systems, and strategic direction through data scattered across many sources.

Core vision

The core idea behind this system is that a university should not be managed through fragmented data kept in files, paper, or disconnected APIs. It should have a data structure that is traceable, reusable, and meaningfully connected to human interpretation.

In the context of Kasetsart University, a system like this can become a conceptual base for data governance and institutional intelligence that does not reduce people to publication counts alone. It should also surface teaching, people development, curriculum building, system building, and the creation of academic ecosystems.

Why the name KU Urban Decision Intelligence

The name is intentionally connected to something the university already had before: KU Forest. In symbolic terms, a forest is a space for gathering resources, diversity, and distributed knowledge. It represents a landscape of data, scholarship, people, outputs, and units growing together within the same ecosystem.

But once a university has a large amount of data, the next question is no longer only, "What do we have in the forest?" It becomes, "When we move into the city, how do we use that data for decision-making?" The city here is not merely physical. It is the space of governance, coordination, tradeoffs, direction-setting, and collective institutional decisions.

For that reason, KU Urban Decision Intelligence reflects a shift from "collecting and storing" toward "interpreting and deciding." If KU Forest is the space of academic resources, KU Urban Decision Intelligence is an attempt to build a layer of intelligence that helps administrators, faculty, researchers, and the university community use those resources together with greater meaning.

The name also signals another idea: a modern university does not only need a data repository. It needs decision infrastructure, meaning a structure that turns data into context, perspective, and strategic conversation that can actually be used.

Data Governance

The goal is not only to pull data onto a web page. It is to ensure that university data sources are handled with provenance, versioning, fallback logic, and clear knowledge of where a piece of data came from, when it was used, and how trustworthy it is.

Human Wisdom

Structured data is never enough to explain the value of a whole person. The system therefore leaves room for personal insight, contextual interpretation, and forms of reading that require human participation rather than allowing metrics to summarize everything by themselves.

Strategic AI

In this system, AI serves as a tool for synthesis, structuring, and pattern recognition at the strategic level. It is not a substitute decision-maker for administrators, experts, or the academic community.

1. Managing data sources and APIs

Many universities already have data, but that data is usually spread across multiple systems, schemas, and units, without a serious canonicalization layer in between. When the institution tries to use data strategically, the same problems repeat: inconsistent naming, inconsistent language, duplicate records, broken continuity, or uncertainty about which field should actually be trusted.

This system therefore experiments with the idea that there should be a clear middle layer between source and presentation, including:

A retrieval layer for APIs and upstream data sources
A transformation layer that turns raw records into a canonical shape shared by the web app and AI
A cache layer that separates latest from last good to survive upstream failures
A metadata layer that explains whether data came from live, cache, fallback, and what limitations apply

If extended at university scale, this structure could grow into data catalogs, schema governance, data quality monitoring, audit trails, and publication policies for institutional data.

2. An approach to profiling

One problem with profiling systems in higher education is that they often direct too much attention toward a narrow set of metrics such as Q1 publications or research funding. As a result, the picture of faculty members, researchers, or units is reduced to ranking logic, even though the real mission of a university is much broader.

This system therefore tries to support profiling that can be read across multiple dimensions, including:

Research and scholarly outputs
Teaching and knowledge transfer
Administrative roles and system building
Curriculum development, people development, and ecosystem building
Context contributed by the profile owner that system data may not yet capture

Profiling here is therefore not about image-making. It is about helping the university see the contribution structure of a person or group more clearly: what kinds of value they create, and during which periods.

3. Using AI to support strategic analysis

The use of AI in this system is based on the principle that AI should help synthesize and open new perspectives rather than issue final judgments. The system therefore prepares prompts from structured data that has already been organized, while clearly separating system intelligence from personal insight contributed by humans.

If developed further at university scale, AI could help in many ways, for example:

Reading trends across disciplines, units, or research groups over time
Synthesizing links among research, teaching, administration, and ecosystem building
Preparing intelligence briefs for policy or executive discussions
Helping leaders ask better questions rather than forcing fixed answers

Even so, AI should never replace human judgment, and it should not be used to normalize incomplete data. The system must therefore always surface limitations, reliability, and context alongside the analysis.

Design principles

Traceable rather than opaque
Interpretable rather than pre-packaged
Bilingual by design for Thai and English use
Capable of fallback when sources are unstable
Open to human-added context rather than closed by the system

If extended at university scale

Connect multiple data sources under a shared governance framework
Build institutional profiles at the levels of individual, department, faculty, and thematic cluster
Use AI to support strategic briefing for leadership
Develop dashboards that show both metric and meaning
Turn the system into decision infrastructure rather than just a single polished web page

Conclusion

The core of this system is an attempt to make data and human wisdom work together. Good data helps reveal pattern and accountability, while people contribute interpretation, value, context, and strategic meaning.

If Kasetsart University wants to move toward more evidence-informed governance without abandoning human understanding, a system like this can serve both as a prototype and as a starting conversation about how data governance, profiling, and AI should serve the university.

Acknowledgements

Provided data and API: Kasetsart University Research and Development Institute (KURDI)

Hosting: Faculty of Engineering server infrastructure