Instructor Notes

This is a placeholder file. Please add content here.

Introduction to Salmon Knowledge Modelling

Reusing Terms — Search and Integrate Existing Vocabularies

Instructor Note

  1. Warm-up Discussion (10 min) Ask:

“What challenges do you face when merging data from other sources?”

“Has anyone tried to interpret someone else’s dataset and gotten confused by a term?” → Summarize: inconsistent naming blocks reuse and synthesis.

  1. Concept: Why Reuse? (10 min) Explain that reusing existing terms ensures that data “speak the same language.” Example:
  1. Demonstration: Searching Existing Vocabularies (15 min) Instructor shares screen:
  • Search for “salmon” or “brood year” on BioPortal.bioontology.org or NVS.
  • Show how to view term metadata (label, definition, URI, license).
  • Demonstrate copying the URI into the Data Dictionary Template.
  1. 🧠 Challenge / Activity 1: Find and Reuse (30 min)

Instructor Note

Group Debrief (10 min) Ask:

  • Which terms were easy to find?
  • Which were hard or missing?
  • When would you decide to reuse vs. define your own?
  1. Reflection (10 min) Discuss the downstream benefits:
  • Reusing terms enables automatic linking and machine-readability.
  • Fewer mapping issues later when integrating salmon datasets.

Documenting Terms — Write Clear, Useful Definitions

Instructor Note

  1. Concept: What Makes a Good Definition? (15 min) Show two examples: ❌ “Run timing: when fish come back.” ✅ “Run timing: The seasonal period during which adult salmon return from the ocean to their natal freshwater spawning areas.” Discuss why clarity, precision, and context matter.

  2. Demonstration: Documenting a Term (10 min) Instructor walks through one dataset column, filling in the template:

Label: brood_year

Definition: “The calendar year when the majority of parental salmon spawned.”

Unit: “Year (YYYY)”

Example: “2017”

Notes: “Equivalent to term in DFO Salmon Concept Scheme.”

  1. Activity: Extract and define (40 min)

Instructor Note

  1. Discussion: Patterns and Pitfalls (15 min)

Which terms were hardest to define?

Were there local abbreviations or codes that need clarification?

How can we document uncertainty? (e.g., “derived from visual estimate”).

  1. Reflection (10 min) Connect to next steps:

A well-documented data dictionary is the foundation for term alignment.

Later modules will link these definitions to others via mappings.


Concept Decomposition

Instructor Note

  1. Welcome & Overview (10 min) Instructor Talking Points

“We’ve documented and clarified what our terms mean — now we’ll start exploring how they relate to each other.”

“This process helps identify overlaps, redundancies, and relationships that will make your vocabulary more consistent and integrable.”

“You’ll learn to see terms not just as words, but as concepts in a network of meaning.”

Key Framing Questions

  • What happens when two teams use different terms for the same thing?
  • How can we make these relationships visible and agreed upon?
  • How might this help us integrate data across programs or agencies?
  • Encourage learners to see this as a detective exercise: uncovering the hidden structure of their vocabulary.
  1. Guided Example: Breaking Down a Concept (10 min) Instructor Demo

Use a familiar term, e.g., “juvenile salmon”, and walk through decomposing it into component ideas.

Ask aloud:

  • What is the core idea here?
  • What attributes or qualifiers are implied (e.g., life stage, species, habitat)?
  • Is this term broader or narrower than another concept we’ve seen?

Discussion Prompt

Ask participants:

“Can you think of a term in your data that’s used differently by different teams or datasets?” Use these examples to highlight why decomposition and mapping relationships improves clarity.

  1. Activity: Decompose and Map Your Own Terms (30 min)

Instructor Tips

Encourage them not to overthink — this is exploratory, not final.

If they get stuck, prompt with:

“What is this a type of?”

“What does this include?”

“What is this related to but not the same as?”

Walk around (or circulate in breakout rooms) and listen for examples that highlight ambiguity or hierarchy.


Instructor Note

  1. Discussion & Wrap-Up (10 min) Group Discussion

Invite volunteers to share their concept maps or a tricky example they encountered.

Ask:

“Where did you find terms that overlap or conflict?”

“How might this process help clarify things for your collaborators?”

“What relationships were hardest to define?”

Encourage connections to real-world integration challenges — e.g., two agencies using different terms for similar stages or metrics.

💡 Instructor Notes & Tips Common Challenges

Learners get bogged down in wordsmithing — remind them that this stage is about structure, not final phrasing.

Difficulty identifying relationships — use analogies (e.g., “species is to genus as narrower is to broader”).

Too narrow a focus — encourage learners to zoom out to see relationships across datasets, not just within one.

Optional Extension

If time allows, show a quick example of a simple schema diagram (e.g., a few boxes and arrows). Explain that this is where they’re heading: transforming their decomposed concepts into a formal structure that can support data interoperability.


From Concepts to Semantics — Introducing SKOS

Instructor Note

Key Takeaways for the Instructor to Reinforce

SKOS is about organizing concepts, not building full ontologies yet.

It’s okay if learners don’t fully “get” RDF — focus on relationships and hierarchy.

Encourage conversation about meaning, consistency, and relationships between concepts.

Diagrams help demystify formal semantics — it’s okay to stay visual!

Facilitator Prompt:

“You’ve all worked on documenting your data terms and even aligning them across datasets. But how do we represent those relationships formally, so others can understand or reuse them — including computers? That’s where SKOS comes in.”

Questions to ask the room:

“What happens if two groups both define ‘condition factor’ slightly differently?”

“How do you think we could show that one term is broader or narrower than another?”

“Why might this matter when sharing data or integrating across studies?”

Instructor Tip:

Keep it conversational — the goal is to surface the problem space that SKOS solves. Don’t introduce jargon yet.

Teaching Flow

  1. Define SKOS in plain language:

“SKOS stands for Simple Knowledge Organization System. It’s a way to represent vocabularies — lists of terms and their relationships — in a structured way computers and humans can understand.”

  1. Relate it to what they already know:

“You already have terms, definitions, and mappings. SKOS gives those structure — think of it as putting your dictionary into a well-organized tree.”

  1. Show an example.

  2. Make the bridge to ontology:

“SKOS is not an ontology — it doesn’t describe processes or logic. But it helps us get there by establishing consistent language.”

Instructor Notes:

If learners seem intimidated, reassure them: “You don’t need to write code today — we’re just organizing concepts visually.”

Have a slide or printed SKOS term table for reference: Concept, prefLabel, definition, broader, narrower, related, exactMatch, closeMatch.


Instructor Note

💬 Reflection (10 min)

Discuss as a group:

  • What patterns or redundancies did you notice in your terms?
  • Which concepts could be reused from an existing vocabulary?
  • How does formalizing these relationships help you answer your Competency Questions from Module 4?

End by connecting this back to Competency Questions (Module 4):

“Your CQs ask big research questions. The SKOS structure helps ensure your vocabulary supports answering those questions consistently.”


From Terms to Meaning - Framing Knowledge with Competency Questions

Instructor Note

  1. Concept Introduction (15 min)

Prompt discussion:

“If your vocabulary or ontology were a smart assistant, what questions would you want it to answer?”

Use this to introduce Competency Questions as a design tool.

Explain: Competency Questions (CQs) help define:

What knowledge must be represented.

How concepts are related.

Which data connections are needed to answer real-world questions.

Example:

“Are differences in sockeye smolt condition at ocean entry contributing to differences in adult return abundances between Fraser River and Bristol Bay populations?” → This implies relationships between smolt condition, population, location, ocean entry timing, and adult return abundance.

Show how each CQ points toward relationships that need to exist in the ontology (e.g., “hasCondition,” “occursAt,” “belongsToPopulation”).

  1. Demonstration: Deriving CQs from Data (10 min)

Instructor uses one term from the data dictionary (e.g., brood_year) and models a possible CQ:

“How does brood year relate to adult return abundance?”

“Which brood years correspond to years of low marine survival?”

Show how these questions hint at:

Entities: brood year, adult return, marine survival rate

Relationships: occurs_in_year, affects, has_rate

Explain that this is the first step in translating vocabulary terms into knowledge relationships.

  1. 🧠 Challenge / Activity 1: Writing Competency Questions (35 min)

Instructor Note

Facilitator Notes:

Encourage specificity. Instead of “What affects salmon survival?”, refine to “Does earlier ocean entry date affect survival of Fraser River Chinook smolts?”

Keep language natural — these aren’t queries yet, just conceptual targets.

  1. 🧩 Challenge / Activity 2: Extracting Relationships (20 min)

Bonus SessionOntology Game Workshop

Instructor Note

Facilitator Tips

  • Keep the tone light and curious — emphasize exploration, not correctness.
  • Encourage laughter when groups disagree — it mirrors real data harmonization.
  • Use salmon-specific examples participants recognize to ground abstraction.
  • End with a call to action: share a link or example ontology relevant to their domain (e.g., EnvO, Darwin Core, OBO Foundry).

Introduction

  • Welcome participants and outline the workshop structure.
  • Introduce the idea that ontologies = shared meaning frameworks — not just fancy dictionaries.
  • Use a relatable example: “If one project measures ‘juvenile abundance’ and another measures ‘smolt density’, do we know if those data can be compared?”

Slide or flipchart talking points:

  • Data integration challenges in salmon research:
    • Different terms for similar things.
    • Ambiguous definitions.
    • Data collected at different scales or life stages.
  • How ontologies help:
    • Clarify relationships between terms.
    • Enable machine-readable data structures.
    • Support data discovery and reuse.

Transition:

“Let’s explore how our own thinking about salmon data reflects some of these challenges.”


Instructor Note

Debrief Discussion (10 min)

Prompt questions:

  • What patterns did you notice?
  • Did your groups agree easily? Where did disagreements arise?
  • Were any terms hard to place?
  • What kinds of groupings emerged — by measurement type, by organism, by process?

Facilitator takeaway:

“This shows how even shared vocabulary can have different implicit structures — an ontology helps make those structures explicit.”


Instructor Note

Miro key:

Spawner abundance | adult salmon population | abundance | census count at weir Capture efficiency | Sampling method event | error | statistical model, such as mark-recapture Ocean-age 3 | population demographic | count | age classification, but how??

Discussion (10 min)

  • Which terms were most complex or ambiguous?
  • How could decomposition help when mapping data between projects?
  • Do you see examples where datasets might use the same word but mean different things?

Facilitator takeaway:

“Ontologies make these hidden structures explicit. When we define relationships clearly, our data become easier to integrate and reason over.”


Instructor Note

Discussion

  • How did your structure clarify meaning?
  • What relationships felt intuitive vs forced?
  • Where did alignment challenges appear when merging?

Facilitator takeaway:

“Ontologies don’t just name concepts — they define how those concepts relate. This structure is what lets systems automatically align datasets.”

Wrap-Up and Reflection (10 min)

Group discussion prompts:

  • Where in your research workflows do you see ambiguity in data terms?
  • Which tasks today reflected real challenges in your projects?
  • What would be the first step towards adopting an ontology in your work?

Facilitator closing:

“By making meaning explicit, ontologies turn local data into shared, interoperable knowledge. Even small steps — like decomposing terms or defining relationships — move us toward more integrated salmon science.”

Optional Gamification Ideas

  • Ontology Bingo: Each participant gets a bingo sheet with boxes like “Found a synonym”, “Identified hidden concept”, “Merged two groups”, “Defined an ‘is-a’ relationship”.
    Mark off during tasks; first to shout “Semantic alignment!” wins a small prize.

  • Ontology Auction: Each group receives a “dataset” card (e.g., spawn timing dataset, tag detections, habitat surveys).
    They can “bid” for ontology terms (like event, location, organism, measurement) to integrate their data — illustrating the value of reusability.