PJQv1.2

PJQ Resonance Assessment Methodology

Transparency and open logic are fundamental to PJQ. Unlike corporate sentiment systems, we publicly document our classification rules as a `rubric.yaml`. The consensus index (Mood-Net) is built only on substantive judgments, denying click-farms and empty likes a way to inflate the numbers.

Consensus Formula

Mood-Net Mathematical Assessment

The consensus index is computed from the Opinion Pool only (SUP / AGA / NEU). The other groups (THIN, OFF, SUS, AGN) are counted separately and have no weight in the resulting support formula:

Mood-Net=
SUP — AGASUP + AGA + NEU
Index range: from -1.00 (full criticism) to +1.00 (full approval)
Category Directory:
SUPSupport
IN FORMULA

Agreement with the author's thesis, scientific or argued defense of the position.

AGAAgainst
IN FORMULA

Constructive disagreement, fact-checking, debunking, counter-arguments.

NEUNeutral
IN FORMULA

Constructive on-topic questions, cautious balanced takes without strong emotion.

THINThin Positive
EXCLUDED

Emotional cliches ('amazing, super'), emoji likes, respect for the blogger without engaging with the substance.

OFFOfftopic
EXCLUDED

Technical questions about audio, timestamps, clothes or clips, unrelated stories.

SUSInorganic
EXCLUDED

Ad spam, recovery scammers, bot farms, click-farming and engagement-bait.

AGNOwn Agenda
EXCLUDED

Commenters siphoning traffic to their own product, Telegram channel or crypto platform.

Filtering Manifest: rubric.yaml

This config lives at the root of the PJQ repository. Any independent researcher can validate our open labeling method by exporting the file directly into their Claude/Gemini client.

version: "1.2"
observatory: "PJQ — Public Judgment Quotient"
classification_rubric:
  SUP:
    name: "Support"
    weight_formula: 1.0
    indicator_keywords: ["agree", "exactly", "deep analysis", "confirmed", "right"]
  AGA:
    name: "Against / Critical"
    weight_formula: -1.0
    indicator_keywords: ["disagree", "mistake", "cliche", "actually not", "debunk"]
  NEU:
    name: "Neutral"
    weight_formula: 0.0
    indicator_keywords: ["on-topic question", "why", "on one hand", "source"]
  THIN:
    name: "Thin Positive"
    influence_on_net: false
    indicator_keywords: ["like", "love", "top", "great", "fire"]
  OFF:
    name: "Offtopic"
    influence_on_net: false
    indicator_keywords: ["audio", "edit", "mic", "music", "jacket"]
  SUS:
    name: "Inorganic / Bot"
    influence_on_net: false
    indicator_keywords: ["airdrop", "telegram", "fast money", "free"]
  AGN:
    name: "Own Agenda"
    influence_on_net: false
    indicator_keywords: ["subscribe to me", "my channel", "doing reviews"]

For Developers: Seismograph API

Run resonance assessment under videos directly from your servers via a simple POST request.

POST/api/v1/classify
{
  "title": "How brain plasticity saves from...",
  "source": "Sci-Outreach",
  "domain": "EDU",
  "sourceTier": "independent",
  "rawComments": "Useful breakdown... \n This is clickbait... \n Great!",
  "runRealAI": true
}