01 // Foundation & Thesis De-Anthropomorphism & Epidemiology

ETHOP-CORE-THESIS / v0.2.0

Overcoming the replication bottleneck in companion animal welfare, born from restless curiosity and empirical rigor.

Ground-level canine clinical intakes and incident archives are shackled to moralizing human prose. The Transparency Project establishes the algorithmic protocols required to convert qualitative narrative noise into immutable, type-strict biotelemetry.

Principal Systems Architect Alice Severi Gonçalves

Persistent Identifier 0009-0003-0048-8982 ↗

Zenodo Software Deposit 10.5281/zenodo.21211371 ↗

Standard Core & Schema Darwin Core (DwC-A v2.4)

CI/CD Suite Verification Python 3.11 & 3.12 (43/43 Passing)

Problem Space // Systematic Noise

The Anthropomorphic Narrative Trap

Inter-rater observer variance exceeds 74% in freehand clinical narratives (Fleiss' κ = 0.22), paralyzing epidemiological discovery.

Qualitative Bottleneck OBSERVATION_ERR // >74%

Shelter intake records, canine shelter logs, and veterinary incident records rely on moralizing, projective vocabulary. Canine subjects are routinely categorized as "spiteful," "guilty," "stubborn," or "mean."

These linguistic shortcuts bypass physiological mechanics. A canine labeled "spitefully destructive" when left alone has its somatic distress flattened. What actually occurred is elevated salivary cortisol, sustained panting, pacing locomotion vectors, and frenzied oral displacement chewing around ingress points.

Inter-Rater Concordance Matrix (Fleiss' Kappa) κ = 0.22 (Poor)

Uncalibrated freeform incident text yields statistically negligible multi-observer consistency across 12 shelter pilot tests.

format_quote

"Human observers routinely substitute emotional projection for motor sequence indexing, transforming empirical science into folkloric fiction."

— EthoPipe Project Manifesto (2025–2026)

Diff Inspector gemini-3.5-flash @ T=0.0

Deterministic Normalization

Input: Raw Volunteer Prose (Uncalibrated) STRING [L=154]

"Dog looked totally guilty after chewing the perimeter fence. When I approached he got stubborn and gave me a mean side-eye, ignoring my commands on purpose."

arrow_downward EthoPipe Clamped Zero-Temp Normalizer (T=0.0)

Output: Darwin Core Star-Schema Observation PARSED JSON [VALID]

{
  "dwc:basisOfRecord": "HumanObservation",
  "dwc:individualID": "CANIS-SUBJ-8821",
  "ethogram_measurements": [
    {
      "dwc:measurementType": "displacement_oral_mastication",
      "substrate": "galvanized_mesh",
      "dwc:measurementValue": 1
    },
    {
      "dwc:measurementType": "displacement_oral_lip_licking",
      "frequency": 4,
      "dwc:measurementUnit": "discrete_count"
    },
    {
      "dwc:measurementType": "pinnae_caudal_retraction",
      "intensity": "grade_2"
    },
    {
      "dwc:measurementType": "locomotion_avoidance_vector",
      "gaze_aversion": true
    }
  ],
  "inferred_anthropomorphisms_stripped": [
    "guilty",
    "stubborn",
    "mean",
    "spiteful"
  ]
}

Schema standard: ISO 8601 / Darwin Core XML / DwC-A check_circle Valid Pydantic v2 Schema

Methodological Architecture

The Three Scientific Pillars of Applied Ethology

How EthoPipe bridges veterinary behavioral field studies with high-velocity algorithmic processing pipelines.

PILLAR 01 Ethogram Matrix

Standardized Kinematic Ethograms

Grounding field notations in peer-reviewed observational frameworks (Broseghini et al., 2024; de Winkel et al., 2024). Every qualitative notation is collapsed into verifiable motor patterns: lip licks, commissure retraction, play bows, caudal wag angles, and weight shifts.

Catalogued Motor Units 148 Distinct Kinematic Keys

PILLAR 02 Clamped Temp 0.0

Clamped Deterministic Parsing Clerks

Utilizing large foundation models (gemini-3.5-flash) stripped completely of creativity. By locking generative temperature strictly to T = 0.0 and applying structured JSON schema decoding, the model functions strictly as an infallible text-cleaning parsing clerk, completely immune to stochastic hallucinations.

Inference Engine gemini-3.5-flash @ T=0.0

PILLAR 03 GBIF / OBIS Standards

Darwin Core Biodiversity Interoperability

Exporting clean observations formatted for the world's primary biological repositories. By mapping animal identifiers to dwc:individualID, events to ISO 8601 dwc:eventDate, behavioral variables to dwc:measurementType, and values to dwc:measurementValue, companion animal behavioral datasets integrate directly into macro-scale veterinary epidemiology.

Relational Topology Star Schema / Parquet / DwC-A

Grounding Biological Literature & Formal Peer Citations

Empirical Foundation Citations

3 Canonical References

Predatory Motor Sequence

Broseghini et al. (2024).

Ethogram of the predatory sequence of dogs (Canis familiaris). Applied Animal Behaviour Science, 279, 106402.

https://doi.org/10.1016/j.applanim.2024.106402 ↗

Canine Welfare Systematic Review

de Winkel et al. (2024).

Observational behaviors and emotions to assess welfare of dogs: A systematic review. Journal of Veterinary Behavior, 72, 1–17.

https://doi.org/10.1016/j.jveb.2023.12.007 ↗

Veterinary Physiological Baselines

Merck Veterinary Manual.

Normal Physiological Values for Dogs: Baseline resting canine heart rates (60–140 BPM), extreme biological telemetry boundaries (30–250 BPM).

Merck Veterinary Reference Compendium

Next Section // 02 Architecture & Docs

Explore the EthoPipe ETL Engine Documentation

Discover how Pydantic type-strict models, clamped Google AI Studio token parsers, and continuous GitHub Actions CI preserve zero-drift reproducibility for veterinary science.

Inspect Pipeline Docs arrow_forward Zenodo Archive open_in_new