Affective Cognition Engine
Emotion is not a facial expression.
It is a cognitive interpretation of reality.
Most emotion AI detects smiles, frowns, or tone. FineSchema analyzes the situation behind the signal: what happened, how the person interpreted it, what expectation was violated, which emotion hypotheses are justified, and what remains uncertain.
감정은 표정이 아니라, 인간이 현실을 해석한 결과입니다.
Recognition structure
From signal to a metacognitive emotion hypothesis.
Input Signals
Text · Voice · Face · Behavior.
Context Layer
Situation · Relationship · History · Social role.
Cognitive Appraisal
Threat · Loss · Rejection · Injustice · Shame · Attachment.
Emotion Hypotheses
Anxiety · Anger · Shame · Guilt · Disappointment · Desire.
Metacognitive Output
Confidence · Uncertainty · Safer response.
Worked example
“I'm fine. Do whatever you want.”
A naive classifier returns ‘Neutral.’ FineSchema reads the situation behind the words.
Scenario
User message
“I'm fine. Do whatever you want.”
Context
Romantic conflict after a delayed response.
Naive AI: Neutral
Press “Analyze Emotion” to run the appraisal.
FineSchema does not claim to read minds. It generates emotional hypotheses with uncertainty and encourages safer human response.
FineSchema는 사람의 마음을 완벽히 읽는다고 주장하지 않습니다. 감정 가설과 불확실성을 함께 제시하고, 더 안전한 대화를 유도합니다.