1     Definition 2     Functional Requirements 3     Syntax
4     Semantics 5    Conformance Testing 6     Performance Assessment

Definition

The digital representation of temporal entities, events, and their relations – observable realisations of events – designed to support semantic annotation of dialogue, text, and multimodal content (according to ISO 24617-2:202).

Functional Requirements

  1. Temporal Semantics – MUST Represent temporal points, intervals, durations, and uncertainty (e.g., bounds, confidence) with normalisation to standard time scales (UTC offsets, calendars).
  2. Event Semantics – MUST Represent events with type, participants/actors, attributes (e.g., modality, polarity, state), and links to time and source media segments.
  3. Expression Semantics – MUST Represent expressions as observable realizations of events in communicative media (linguistic forms, gestures, vocalisations, multimodal composites), including structure, content, and association to the event(s) they realize.
  4. Expression–Event Linking – MUST Provide explicit links from expressions to the event(s) and time(s) they realise, supporting one‑to‑one, one‑to‑many, and many‑to‑one
  5. Expression Typing – MUST Support typing of expressions (e.g., utterance, token span, gesture segment, vocal cue, composite) and sub‑typing via controlled vocabularies.
  6. Multimodal Anchoring – MUST Provide anchoring mechanisms to reference spans/segments in text, audio, video, images, or sensor data (e.g., offsets, timestamps, frame indices) for both expressions and events.
  7. Identity and Referencing – MUST Ensure stable identifiers for Time, Event, and Expression entities, cross‑reference capability, and uniqueness constraints within an annotation instance.
  8. Trace – MUST Record provenance metadata (origin, timestamp, tool/user, version) for each entity and change operation.
  9. Extensibility – MUST Allow controlled extension points (namespaces, custom attributes, vocabularies) without breaking core conformance.
  10. Constraints and Profiles – SHOULD Support domain‑specific profiles (e.g., dialogue acts, clinical discourse, media production) with additional constraints and enumerations.
  11. Quality Measures – SHOULD Provide optional confidence scores, quality flags, and validation reports for expressions, events, and links.
  12. Queryability – SHOULD Facilitate efficient querying by time window, event type, expression type, participant, or modality.

Syntax

https://schemas.mpai.community/MMC/V2.5/data/TimeEventExpression.json

Semantics

Label Description
Header Time Event Expression Header
Standard‑TEE The characters ‘MMC‑TEE‑V’ marking the Time-Event-Expression data type.
Version Major version – 1 or 2 characters.
Dot‑separator The character ‘.’ separating version components.
Subversion Minor version – 1 or 2 characters.
TEEAnnotationID Unique identifier for this TEE annotation instance.
Time Set of Time entities representing temporal points, intervals, or durations.
participants Array of participant objects; items are unique.
– id String identifier of the participant.
– role Role of the participant: speaker | addressee | audience | system | other.
– type Type of participant: human | agent | system.
– label Human-readable label for the participant.
– metadata Participant-specific metadata.
functionalSegments Array of functional segment objects.
– id String identifier of the functional segment.
– source Reference to the primary data resource (e.g., document URI, audio/video file).
– modality Segment modality: text | speech | gesture | multimodal | other.
– language Language code for the segment.
– start Numeric start position/time.
– end Numeric end position/time.
– offsetType Offset units: char | token | ms | frame | sample.
– text Text content of the segment (if applicable).
communicativeFunctions Array registry of communicative functions (may reference an external vocabulary).
– id String identifier of the communicative function.
– label Human-readable label of the communicative function.
– dimension Dimension: Task | Feedback | TurnManagement | TimeManagement | DiscourseStructuring | OwnCommunicationManagement | PartnerCommunicationManagement | SocialObligationsManagement.
– definition Definition text of the function.
– uri URI of the function entry (if any).
dialogueActs Array of dialogue act objects.
– id String identifier of the dialogue act.
– segmentRef Reference to a functionalSegments[id].
– senderRef Reference to the sender participants[id].
– addresseeRefs Array of references to participants[id] for addressees.
– functionRef Reference to communicativeFunctions[id].
– dimension Dimension of the act; same enumeration as communicativeFunctions.dimension.
– semanticContent Optional object container for domain semantics.
– qualifiers.sentiment String sentiment qualifier.
– qualifiers.certainty Certainty value in range 0.0–1.0.
– qualifiers.conditional Boolean conditional flag.
– dependence.functional DialogueAct[id] this act depends on (e.g., answer depends on question).
– dependence.feedback DialogueAct[id] this act gives feedback to.
– rhetoricalRelations Array of rhetorical relation objects attached to the act.
  – type Type of rhetorical relation (string).
  – targetActRef Target DialogueAct reference (string).
– trace
  – origin Origin string for provenance.
  – timestamp Timestamp in date-time format.
  – tool Tool identifier string.
  – version Version string.
relations Array of non-dependence relations among dialogue acts.
– id String identifier of the relation.
– sourceActRef Source DialogueAct reference (string).
– targetActRef Target DialogueAct reference (string).
– relationType Type of relation (string).
– confidence Confidence value in range 0.0–1.0.
– trace.origin Origin string for provenance.
– trace.timestamp Timestamp in date-time format.

5     Conformance Testing

A Data instance Conforms with MPAI-MMC Time Event Expression (MMC-TEE) if:

  1.  Its JSON Object validates against its JSON Schema.
  2. Any included  JSON Object validates against its JSON Schema.
  3. All Data in the JSON Object:
    1. Have the specified Data Types.
    2. Conform with the Qualifiers signaled in their JSON Schemas.

6     Performance Assessment