From Data to Digital-Opportunity Intelligence
A hands-on learning repository for building fluency across relational query languages, logic programming, semantic knowledge graphs, property graphs, graph traversal, path querying, and network science.
This repository accompanies my learning roadmap for SQL, first-order logic, Prolog, Datalog, RDF, SPARQL, Cypher, GQL, Gremlin, graph path querying, and network science. Rather than treating each language as an isolated technology, the project uses one continuing domain model: a small Michigan Digital Opportunity Knowledge Graph.
The objective is not simply to learn syntax. It is to understand how the same real-world system looks when represented as tables, logical statements, semantic triples, property graphs, traversals, paths, and networks, and to recognize which representation fits the question being asked.
Why This Repository Exists
My work with MiGreatDataLake, the Digital Opportunities Intelligence Network (DOIN), knowledge graphs, community data, and research increasingly sits at the intersection of data engineering, semantics, network science, and causal inquiry. That makes understanding query languages more than a database skill. It is part of understanding how information becomes knowledge and how relationships become analyzable structures.
The long-term progression behind this repository is:
Data
↓
Relationships
↓
Semantics
↓
Knowledge
↓
Graph Structure
↓
Research Questions
↓
Statistical Models
↓
Causal Inquiry
A knowledge graph is therefore not the endpoint. It is a connective analytical layer between data and increasingly sophisticated forms of inquiry.
Learning Roadmap
| Stage | Topic | Primary Goal |
|---|---|---|
| 1 | Sets, Relations & Relational Algebra | Understand the mathematical foundations of relational data |
| 2 | SQL | Become fluent in relational querying and recursive queries |
| 3 | First-Order Logic | Distinguish facts, predicates, rules, and hypotheses |
| 4 | Prolog | Make rule-based reasoning executable |
| 5 | Datalog | Learn recursive and deductive database reasoning |
| 6 | RDF | Model information as semantic subject-predicate-object statements |
| 7 | SPARQL | Query and construct RDF graph patterns |
| 8 | Property Graphs & Cypher | Develop graph-pattern thinking |
| 9 | GQL | Understand standardized property-graph concepts |
| 10 | Gremlin | Develop traversal-oriented graph reasoning |
| 11 | Path Querying | Study reachability, variable-length paths, cycles, and navigation |
| 12 | Network Science | Analyze structure rather than merely retrieve relationships |
The complete curriculum, exercises, capstone questions, and learning rationale are in query-language-learning-roadmap.md.
One Domain, Many Representations
The central learning strategy is simple: keep the domain relatively stable while changing the information model and query paradigm.
The working domain may include:
- Communities and counties
- Schools and intermediate school districts
- Libraries
- Nonprofits
- Tribal Nations
- Broadband providers
- Workforce organizations
- State agencies
- Universities and research organizations
- Programs and initiatives
- Funding sources
- Infrastructure projects
Representative relationships include:
Community → LOCATED_IN → County
Organization → SERVES → Community
Organization → PARTNERS_WITH → Organization
Organization → PARTICIPATES_IN → Program
ISP → PROVIDES_SERVICE_TO → Community
Agency → FUNDS → Program
Program → SUPPORTS → Community
The exact model will evolve as the project develops. Changes to the model should be intentional and documented so that differences among representations remain part of the learning record.
Repository Structure
The intended repository structure is:
.
├── README.md
├── query-language-learning-roadmap.md
├── data/
│ ├── raw/
│ ├── processed/
│ └── sample/
├── 01-relational-algebra/
│ └── notes.md
├── 02-sql/
│ ├── schema.sql
│ ├── seed.sql
│ ├── queries.sql
│ └── journal.md
├── 03-first-order-logic/
├── 04-prolog/
├── 05-datalog/
├── 06-rdf/
│ ├── ontology/
│ └── data/
├── 07-sparql/
├── 08-cypher/
├── 09-gql/
├── 10-gremlin/
├── 11-path-querying/
├── 12-network-science/
├── capstone/
└── docs/
This is the target organization for the learning repository, not a claim that every directory or artifact already exists.
What Each Stage Should Produce
Each stage should preserve four core learning artifacts.
1. Data Model
Document how the digital-opportunity ecosystem is represented at that stage. A relational schema, RDF vocabulary, property-graph model, or rule set is itself part of the research record.
2. Working Queries or Rules
Maintain at least a small progression from simple retrieval to multi-hop or recursive questions. The emphasis should be on understanding what each expression means, not accumulating query files.
3. Query Journal
For important queries, explain in plain English:
- What question am I asking?
- What information does the query rely on?
- What assumptions are embedded in it?
- What does the result establish?
- What does the result not establish?
This is especially important as the project moves from retrieval into inferred relationships and network analysis.
4. Comparison Note
At the end of each stage, document what became easier, harder, or conceptually different compared with the previous representation.
For example, the valuable lesson in rewriting an SQL query in SPARQL or Cypher is not merely learning different punctuation. The comparison should reveal how relational joins, RDF graph patterns, property-graph patterns, and traversals express relationships differently.
Core Query Ladder
The repository should move through increasingly sophisticated classes of questions.
Descriptive
What organizations operate in this county?
Relational
Which organizations serve the same communities?
Semantic
Which entities are connected through explicitly defined types of relationships?
Path
How is a community connected to a state program in the modeled data?
Deductive
Under explicitly defined rules, which communities could be connected to workforce resources through partner organizations?
Structural
Where does the represented network contain clusters, bridges, or disconnected components?
Comparative
How does the observed institutional network surrounding rural communities differ structurally from the network represented around urban communities?
Statistical
Is observed institutional network connectivity associated with measured digital-opportunity outcomes?
Causal
Under an appropriate research design, is there evidence that a particular intervention or network condition causes a change in an outcome?
The last two questions require methods beyond graph querying alone. Network structure can generate hypotheses and explanatory variables, but association and graph position should not be treated automatically as causal evidence.
Capstone: DOIN Digital-Opportunity Knowledge Graph
The capstone is a bounded, inspectable representation of a real digital-opportunity ecosystem. A practical learning target is roughly 100–250 entities, large enough to create meaningful paths and structure while remaining small enough to inspect during the learning process. This is a project-design choice rather than a technical requirement.
Potential entity classes include:
People
Organizations
Communities
Counties
Schools
Libraries
ISPs
Tribal Nations
State Agencies
Universities
Nonprofits
Programs
Funding Sources
Infrastructure
Research Projects
Potential relationships include:
SERVES
LOCATED_IN
FUNDS
PARTNERS_WITH
EMPLOYS
RESEARCHES
OPERATES
PARTICIPATES_IN
PROVIDES_SERVICE_TO
SUPPORTS
The capstone is successful when I can ask the same substantive question through multiple representations, explain what each representation contributes, and distinguish stored facts from inferred relationships, structural measurements, hypotheses, and causal claims.
Learning Priorities
For my current work, I am prioritizing the following areas:
| Priority | Topic | Target |
|---|---|---|
| ★★★★★ | SQL | Fluent |
| ★★★★★ | RDF/SPARQL | Fluent |
| ★★★★★ | Cypher | Fluent |
| ★★★★★ | Path and pattern querying | Deep conceptual understanding |
| ★★★★★ | Network science | Strong |
| ★★★★☆ | Datalog | Conceptually strong |
| ★★★★☆ | GQL | Working knowledge |
| ★★★☆☆ | Gremlin | Working competency |
| ★★★☆☆ | First-order logic | Conceptually strong |
| ★★☆☆☆ | Prolog | Enough to understand logic programming |
These priorities reflect the intended application to MiGreatDataLake, DOIN, knowledge graphs, network analysis, and research. They are not intended as a universal ranking of the technologies.
Provenance and Research Discipline
A major principle of this repository is maintaining clear provenance as the work becomes more analytical. Every result should be distinguishable as one of the following:
- Source data: an observation or value obtained from an identified source.
- Modeled relationship: a relationship explicitly represented from source evidence.
- Derived result: something produced through a documented query or rule.
- Network measure: a computed property of the modeled graph.
- Interpretation: a reasoned explanation of what a result may mean.
- Hypothesis: an empirical proposition requiring testing.
- Causal claim: a claim requiring an appropriate identification strategy and supporting evidence.
This distinction matters because a knowledge graph can make relationships visible without proving why those relationships exist, whether they are complete, or whether they cause observed outcomes.
Progression at a Glance
Relational Algebra
│
SQL
│
First-Order Logic
│
┌────┴────┐
Prolog Datalog
│
RDF
│
SPARQL
│
┌─────────┴─────────┐
│ │
Cypher Gremlin
│
GQL
│
Path Querying
│
Graph Algorithms
│
Network Science
│
Statistical Models
│
Causal Inference
Guiding Principle
The measure of progress is not how many languages I can list. It is whether I can recognize the nature of the question in front of me.
Is it relational? Semantic? Deductive? Structural? Navigational? Statistical? Causal?
That distinction determines what data model, query language, analytical method, and evidence are appropriate. The repository is intended to make that progression visible through working examples rather than leave it as an abstract framework.
About
This learning project is developed by Jason F. Kronemeyer as part of a broader exploration of community technology, digital opportunity, intelligent infrastructure, knowledge graphs, network science, research-to-practice translation, and data-informed public problem solving.