6 Relationship Definitions

Relationship Definitions

1. Person -> Meeting

   • ATTENDED: Indicates a person was present at a meeting
   • Direction: Person to Meeting
   • Example: (Person {name: 'John'}) -[:ATTENDED]-> (Meeting {date: '2025-01-01'})

2. Meeting -> ActionItem

   • HAS_ACTION: Links a meeting to tasks or actions decided during the meeting
   • Direction: Meeting to ActionItem
   • Example: (Meeting) -[:HAS_ACTION]-> (ActionItem {text: 'Review proposal'})

3. ActionItem -> Person

   • ASSIGNED_TO: Shows who is responsible for completing an action item
   • Direction: ActionItem to Person
   • Example: (ActionItem {text: 'Review'}) -[:ASSIGNED_TO]-> (Person {name: 'Jane'})

4. Meeting -> Document

   • HAS_DOCUMENT: Links a meeting to its related documents
   • Direction: Meeting to Document
   • Example: (Meeting) -[:HAS_DOCUMENT]-> (Document {title: 'Proposal v1'})

5. Meeting -> Decision

   • MADE_DECISION: Connects a meeting to decisions made during it
   • Direction: Meeting to Decision
   • Example: (Meeting) -[:MADE_DECISION]-> (Decision {decision: 'Approved proposal'})

6. Meeting -> Topic

   • COVERS_TOPIC: Shows topics discussed in a meeting
   • Direction: Meeting to Topic
   • Example: (Meeting) -[:COVERS_TOPIC]-> (Topic {name: 'Governance'})

7. Meeting -> Emotion

   • HAS_EMOTION: Indicates the emotional context or tone of a meeting
   • Direction: Meeting to Emotion
   • Example: (Meeting) -[:HAS_EMOTION]-> (Emotion {name: 'Productive'})

Key Points:

• All relationships are directed (one-way)
• Meetings are the central node type, connecting to most other nodes
• Action Items create a chain from Meeting through Action to Person
• Topics and Emotions provide context and categorization
• Documents track meeting materials and references

Graph Metrics Relationships

// 1. Degree and Path Length
MATCH (n)
OPTIONAL MATCH (n)-[r]-(neighbor)
WITH n, 
     count(DISTINCT neighbor) as degree,
     labels(n)[0] as nodeType
MATCH path = (n)-[*1..3]-(connected)
WITH n, nodeType, degree, connected, path
RETURN 
    nodeType,
    CASE nodeType
        WHEN 'Person' THEN n.name
        WHEN 'Meeting' THEN n.workgroup
        ELSE coalesce(n.title, n.text, '')
    END as nodeName,
    degree as direct_connections,
    count(DISTINCT connected) as reachable_nodes,
    avg(length(path)) as avg_path_length
ORDER BY degree DESC
LIMIT 10;

// 2. Degree Distribution
MATCH (n)
OPTIONAL MATCH (n)-[r]-()
WITH labels(n)[0] as nodeType,
     count(DISTINCT r) as degree
RETURN 
    nodeType,
    degree as connectivity,
    count(*) as node_count
ORDER BY nodeType, degree;

// 3. Clustering Analysis
MATCH (n)-[r1]-(neighbor)
WITH n, 
     collect(DISTINCT neighbor) as neighbors,
     count(DISTINCT neighbor) as degree
WHERE degree > 1
MATCH (n1)-[r2]-(n2)
WHERE n1 IN neighbors 
AND n2 IN neighbors
AND id(n1) < id(n2)
RETURN 
    labels(n)[0] as nodeType,
    degree as node_degree,
    count(*) as neighbor_connections,
    (2.0 * count(*)) / (degree * (degree - 1)) as clustering_coefficient
ORDER BY clustering_coefficient DESC;