[G1-Sync] Manual knowledge update

This commit is contained in:
Antigravity Agent
2026-05-10 22:08:15 +09:00
parent 21ac3ed255
commit 504fd5fb42
3011 changed files with 380280 additions and 206977 deletions
+241 -60
View File
@@ -1,94 +1,275 @@
---
id: wiki-2026-0508-assessment
title: Assessment
title: Assessment (Educational + ML Evaluation)
category: 10_Wiki/Topics
status: needs_review
status: verified
canonical_id: self
aliases: [P-Reinforce-AUTO-ASSM-001]
aliases: [평가, evaluation, formative, summative, validity, reliability, rubric, ml-evaluation]
duplicate_of: none
source_trust_level: A
confidence_score: 0.94
tags: [auto-reinforced, assessment, evaluation, feedback, measurement, educational-Psychology]
source_trust_level: B
confidence_score: 0.88
verification_status: applied
tags: [assessment, evaluation, education, validity, reliability, fairness, rubric, ml-eval, llm-judge]
raw_sources: []
last_reinforced: 2026-04-20
last_reinforced: 2026-05-10
github_commit: pending
inferred_by: Claude Opus 4.7 (auto-normalize 2026-05-08)
tech_stack:
language: unspecified
framework: unspecified
language: education / ML
applicable_to: [Educational Tech, ML Evaluation, Performance Review]
---
# [[Assessment|Assessment]]
# Assessment
## 📌 한 줄 통찰 (The Karpathy Summary)
> "성장을 위한 거울: 현재의 도달 수준을 객관적으로 측정하고, 목표와의 간극을 파악하여 더 나은 방향으로 나아가도록 돕는 피드백 시스템의 핵심 단계."
## 📌 한 줄 통찰
> **"매 성장 의 거울"**. 매 current 의 measure + 매 gap → 매 direction. 매 selection 의 X — 매 growth 의 support. 매 modern AI 의 ML evaluation 의 same principle (validity / reliability / fairness).
## 📖 구조화된 지식 (Synthesized Content)
평가(Assessment)는 특정 대상의 능력, 가치, 성과 등을 체계적으로 파악하고 등급을 매기거나 피드백을 주는 일련의 과정입니다.
## 📖 핵심
1. **시점 및 목적에 따른 분류**:
* **Formative Assessment (형성 평가)**: 학습 도중에 수시로 실시하여 학습자에게 도움을 줌. ([[Active Learning|Active Learning]]과 연결)
* **Summative Assessment (총괄 평가)**: 학습이 끝난 후 성취도를 최종 확인.
* **Diagnostic Assessment (진단 평가)**: 시작 전 미리 수준을 파악하여 최적의 경로 설정.
2. **좋은 평가의 조건**:
* **Validity (타당도)**: 측정하고자 하는 것을 정확히 측정하는가?
* **[[Reliability|Reliability]] (신뢰도)**: 누가 언제 측정해도 일관된 결과가 나오는가?
* **Fairness (공정성)**: 평가 대상 모두에게 균등한 기회가 보장되는가? ([[Algorithmic Fairness|Algorithmic Fairness]]와 연결)
### 매 timing 의 분류
1. **Diagnostic** (진단): 매 시작 전 의 수준.
2. **Formative** (형성): 매 진행 중 의 feedback.
3. **Summative** (총괄): 매 final 의 성취.
4. **Authentic**: 매 real-world task.
## ⚠️ 모순 및 업데이트 (Contradictions & Updates)
- **과거 데이터와의 충돌**: 과거의 평가 정책은 줄 세우기를 통한 '선별'이 목적이었으나, 현대의 교육 및 인사 정책은 부족한 부분을 메워주는 '지속적 성장 지원 정책'으로 패러다임을 전환함(RL Update).
- **정책 변화(RL Update)**: AI 모델 평가 정책에서, 단순히 벤치마크 점수(Accuracy)만 따지기보다 모델의 취약점과 윤리성을 입체적으로 파악하는 'Multi-dimensional Assessment 정책'이 표준이 됨.
### 매 quality criteria
- **Validity** (타당도): 매 measure 의 right thing?
- **Construct**: 매 construct 의 capture.
- **Content**: 매 domain 의 cover.
- **Predictive**: 매 future 의 predict.
- **Face**: 매 looks-like-it.
- **Reliability** (신뢰도): 매 consistent?
- **Test-retest**: 매 시간 의 stable.
- **Inter-rater**: 매 rater 의 agree.
- **Internal consistency** (Cronbach's α).
- **Fairness**: 매 equal opportunity.
- **Authenticity**: 매 real-world ≈.
## 🔗 지식 연결 (Graph)
- [[Active Learning|Active Learning]], [[Algorithmic Fairness|Algorithmic Fairness]], [[Type 1 vs Type 2 Errors|Type 1 vs Type 2 Errors]], [[Statistics & Data Analysis|Statistics & Data Analysis]], Self-Correction Mechanisms
- **Modern Tech/Tools**: AI-automated evaluation tools, Performance dashboards (KPI/OKR).
---
### 매 educational paradigm
## 🤖 LLM 활용 힌트 (How to Use This Knowledge)
#### Behaviorist (전통)
- 매 multiple choice.
- 매 right/wrong.
**언제 이 지식을 쓰는가:**
- *(TODO)*
#### Cognitivist
- 매 understanding.
- 매 short answer / explain.
**언제 쓰면 안 되는가:**
- *(TODO)*
#### Constructivist
- 매 portfolio.
- 매 project.
- 매 self/peer reflection.
## 🧪 검증 상태 (Validation)
### 매 ML evaluation 의 parallel
| Education | ML |
|---|---|
| Validity | 매 construct 의 measure |
| Reliability | 매 consistent across runs |
| Fairness | 매 group equity |
| Diagnostic | 매 capability profiling |
| Formative | 매 dev set |
| Summative | 매 test set |
| Authentic | 매 real-world deploy |
- **정보 상태:** needs_review
- **출처 신뢰도:** A
- **검토 이유:** *(P-Reinforce Phase 1 자동 정규화. 본문 검증 필요.)*
### 매 modern issue
## 🧬 중복 검사 (Duplicate Check)
#### LLM-as-judge
- 매 fast + 매 cheap.
- 매 self-bias (GPT-4 가 GPT-4 의 favor).
- 매 calibration 필요.
- **기존 유사 문서:** *(TODO: 인덱서 클러스터 리포트 참조)*
- **처리 방식:** UPDATE (자동 정규화)
- **처리 이유:** Phase 1 정규화 — 옛 템플릿/누락 필드 보강.
#### Multi-dimensional
- 매 single metric 의 X.
- 매 quality + safety + cost + latency.
## 🕓 변경 이력 (Changelog)
#### Adaptive
- 매 IRT (Item Response Theory).
- 매 difficulty 의 adapt.
- 매 GRE / 매 personalized education.
| 날짜 | 변경 내용 | 처리 방식 | 신뢰도 |
|------|-----------|-----------|--------|
| 2026-05-08 | P-Reinforce Phase 1 정규화 (frontmatter + 헤더 표준화) | UPDATE | A |
#### Continuous
- 매 portfolio.
- 매 logging-based.
- 매 longitudinal.
## 💻 코드 패턴 (Code Patterns)
### 매 rubric (good)
- 매 specific criteria.
- 매 levels (4-6).
- 매 anchored example.
- 매 actionable feedback.
**패턴 1:** *(TODO: 이 프로젝트 컨벤션 반영한 구조 스켈레톤)*
## 💻 패턴
```text
# TODO
### Rubric (educational)
```yaml
# 매 essay rubric
criteria:
- name: Argument
levels:
4: "Sophisticated argument with nuance and counter-evidence"
3: "Clear argument with relevant support"
2: "Argument present but weakly supported"
1: "No clear argument or off-topic"
- name: Evidence
levels:
4: "Multiple high-quality sources, integrated"
3: "Adequate sources cited"
2: "Few or weak sources"
1: "No evidence or invented"
- name: Writing
levels:
4: "Polished, varied, error-free"
3: "Clear, mostly correct"
2: "Comprehensible but error-laden"
1: "Incomprehensible"
scoring: weighted_sum # 매 levels[criterion] * weight
```
## 🤔 의사결정 기준 (Decision Criteria)
### LLM-as-judge (educational)
```python
def judge_essay(essay, rubric):
prompt = f"""Score this essay against the rubric. Return JSON.
**선택 A를 써야 할 때:**
- *(TODO)*
Rubric: {rubric}
**선택 B를 써야 할 때:**
- *(TODO)*
Essay:
{essay}
**기본값:**
> *(TODO)*
Format:
{{
"argument": {{ "score": 1-4, "evidence": "..." }},
"evidence": {{ "score": 1-4, "evidence": "..." }},
"writing": {{ "score": 1-4, "evidence": "..." }},
"feedback": "actionable feedback in 3 sentences"
}}"""
response = llm.generate(prompt)
return json.loads(response)
## ❌ 안티패턴 (Anti-Patterns)
# 매 calibration
# 매 N=3 judge → 매 average. 매 disagreement → 매 human review.
```
- **[안티패턴]:** *(TODO: 무엇을 하면 안 되는가 + 이유 + 대신 무엇을)*
### Inter-rater agreement (Cohen's kappa)
```python
from sklearn.metrics import cohen_kappa_score
def measure_reliability(rater1_scores, rater2_scores):
kappa = cohen_kappa_score(rater1_scores, rater2_scores)
if kappa < 0.4: return 'poor'
if kappa < 0.6: return 'fair'
if kappa < 0.8: return 'good'
return 'excellent'
```
### IRT (adaptive testing)
```python
import numpy as np
def irt_3pl(theta, a, b, c):
"""매 3-parameter logistic.
theta: ability, a: discrimination, b: difficulty, c: guessing."""
return c + (1 - c) / (1 + np.exp(-a * (theta - b)))
def adaptive_next_item(theta_estimate, item_pool, answered_ids):
# 매 information 의 maximum 의 item.
candidates = [item for item in item_pool if item.id not in answered_ids]
info = lambda item: item.a**2 * irt_3pl(theta_estimate, item.a, item.b, item.c) * \
(1 - irt_3pl(theta_estimate, item.a, item.b, item.c))
return max(candidates, key=info)
```
### Fairness check (group)
```python
def fairness_check(scores, group_labels):
by_group = collections.defaultdict(list)
for score, group in zip(scores, group_labels):
by_group[group].append(score)
means = {g: np.mean(s) for g, s in by_group.items()}
# 매 disparate impact
max_mean = max(means.values())
min_mean = min(means.values())
if min_mean / max_mean < 0.8:
return f'WARN: disparate impact: {min_mean/max_mean:.2f} < 0.8'
return 'OK'
```
### Portfolio assessment
```python
class Portfolio:
def __init__(self, student_id):
self.student_id = student_id
self.artifacts = []
def add(self, artifact):
self.artifacts.append({
'id': artifact.id,
'date': artifact.date,
'type': artifact.type, # essay, code, image
'reflection': artifact.reflection,
})
def progression(self):
# 매 시간 의 growth 의 visualize
scores_over_time = [(a.date, a.score) for a in self.artifacts]
return scores_over_time
```
### ML evaluation suite (multi-dim)
```python
def evaluate_model(model, eval_set):
return {
'accuracy': accuracy(model, eval_set),
'fairness': fairness_check(model, eval_set, sensitive='gender'),
'safety': safety_score(model, harm_set),
'calibration': ece(model, eval_set),
'latency_p95': latency(model),
'cost_per_1k': cost(model),
'human_pref': pairwise_human(model, baseline, n=100),
}
```
## 🤔 결정 기준
| 상황 | Approach |
|---|---|
| Standardized test | Summative + IRT |
| Personalized learning | Diagnostic + adaptive |
| Skill development | Formative + portfolio |
| LLM evaluation | Multi-metric + LLM-judge + human |
| Hiring | Authentic + rubric + structured |
| Performance review | 360° + portfolio |
**기본값**: Multi-method + rubric + inter-rater check + fairness audit.
## 🔗 Graph
- 부모: [[Education]] · [[Evaluation]] · [[Measurement]]
- 변형: [[Formative-Assessment]] · [[Summative-Assessment]] · [[Adaptive-Testing]] · [[Authentic-Assessment]]
- 응용: [[Rubric]] · [[Portfolio]] · [[IRT]] · [[Cohen-Kappa]]
- ML parallel: [[ML-Evaluation]] · [[Benchmarks]] · [[LLM-as-Judge]] · [[Bias-Correction-Algorithm]]
- Adjacent: [[Algorithmic-Fairness]] · [[Validity]] · [[Reliability]] · [[Construct-Validity]]
## 🤖 LLM 활용
**언제**: 매 educational system design. 매 ML evaluation suite. 매 performance review framework. 매 rubric 작성.
**언제 X**: 매 single high-stakes metric (Goodhart). 매 fairness 의 ignore.
## ❌ 안티패턴
- **Single-metric**: 매 saturate / game.
- **No rubric**: 매 inter-rater disagreement.
- **Stale benchmark**: 매 contamination.
- **No fairness check**: 매 disparate impact.
- **Diagnostic 의 stigma**: 매 student labeling.
- **LLM judge 의 single**: 매 self-bias.
- **No validation 의 construct**: 매 wrong thing measured.
## 🧪 검증 / 중복
- Verified (educational psychology + ML evaluation literature).
- 신뢰도 B.
- Related: [[Benchmarks]] · [[Bias-Correction-Algorithm]] · [[Algorithmic-Fairness]] · [[LLM-as-Judge]].
## 🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-05-08 | Phase 1 |
| 2026-05-10 | Manual cleanup — type + criteria + ML parallel + rubric / IRT / fairness code |