Skip to content

How to Create Anchor Terms

This guide explains how to create and manage anchor terms for tracking semantic evolution.

What Are Anchor Terms?

Anchor terms are key concepts to be tracked across historical periods. They serve as current/modern reference points or "mile markers" for semantic change analysis.

Examples: - "agent" in AI/philosophy contexts - "machine" in industrial/computing contexts - "intelligence" across cognitive science literature

Creating an Anchor Term

OntExtract offers two ways to create anchor terms: using Connected Services (recommended) or Manual Entry.

Use authoritative lexicographic sources to establish well-documented anchor terms.

Add Term - Connected Services

  1. Navigate to Library → Terms and click Add New Anchor Term
  2. Select a service to search:
Service Description
Merriam-Webster Dictionary Authoritative American English definitions
Merriam-Webster Thesaurus Synonyms and antonyms for context anchoring
Oxford English Dictionary Authoritative English language definitions
WordNet Lexical database with semantic relationships
  1. Enter the search term when prompted
  2. Browse the results and click Use This Definition
  3. The form auto-populates with:
  4. Term text
  5. Meaning description
  6. Context anchors (related terms)
  7. Source citation
  8. Add a Research Domain if desired
  9. Click Create Term

Method 2: Manual Entry

For terms not found in dictionaries or from custom sources:

  1. Click the Manual Entry tab
  2. Fill in the term details:
Field Description Example
Term Text The word or phrase to track "agent"
Source Type Type of reference source Corpus, Dictionary, Standard
Meaning Description Current baseline meaning "An entity that acts autonomously..."
Context Anchors Related terms (comma-separated) "autonomous, actor, entity"
Corpus Source Lexicographic source name "Merriam-Webster"
Source Citation Full citation for provenance "Merriam-Webster Dictionary, 2024"
Research Domain Subject area "Artificial Intelligence"
Notes Additional context "Focus on autonomous agent concept"
  1. Click Create Term

Context Anchors

Context anchors are related terms that help define the semantic space around an anchor term. Options include:

  • Letting services auto-populate them from definitions
  • Clicking the Thesaurus button to find synonyms
  • Entering them manually (comma-separated)

Term Versions

Anchor terms support temporal versioning to capture meaning changes over time. Versions are created through:

  • OED Timeline - Select historical waypoints showing when new senses emerged
  • Manual versioning - Add versions with different temporal periods

Planned Feature

OED Timeline visualization for selecting historical waypoints is under consideration for a future release.

Best Practices

Choosing Terms

  • Specificity: Choose terms specific enough to track meaningfully
  • Frequency: Select terms that appear across the document corpus
  • Evolution potential: Pick terms likely to show semantic change

Term Naming

  • Use the canonical/base form (e.g., "agent" not "agents")
  • Be consistent with capitalization
  • Include phrases if tracking multi-word concepts ("artificial intelligence")

Domain Classification

Assign appropriate domains to help organize terms: - Philosophy - Computer Science - Artificial Intelligence - Cognitive Science - (Custom domains as needed)

Viewing Term Details

Click any term to see:

  • Basic Information - Term text, meaning description, research domain
  • Source Information - Corpus source and citation
  • Context Anchors - Related terms for semantic anchoring
  • Temporal Versions - Historical meaning snapshots (if created)
  • Associated Experiments - Where the term is used

Term Detail Page

Editing Terms

  1. Navigate to the term detail page
  2. Click Edit Term
  3. Modify fields as needed
  4. Click Save Changes

Note: Editing a term affects all experiments using it.

Deleting Terms

To delete a term:

  1. Navigate to Library → Terms
  2. Find the term in the list
  3. Click the trash icon on the right side of the row
  4. Confirm the deletion in the modal

Warning: Deleting a term removes all temporal versions and associated data permanently.

Find existing terms using:

  • Search box - Search by term text
  • Domain filter - Filter by research domain