Educational MCP Server

"""情绪分析模块（规则占位，可换真实模型）。""" from __future__ import annotations from typing import Dict import database from schemas import AffectiveAnalysisRequest, AffectiveAnalysisResponse, AffectiveState, SentimentRequest, SentimentResponse POSITIVE_WORDS = {"好", "满意", "喜欢", "清晰", "有趣", "赞", "棒"} NEGATIVE_WORDS = {"差", "糟", "难", "晦涩", "失望", "生气", "不满", "不会", "不懂", "好难", "不知道"} def _simple_score(text: str) -> Dict[str, float]: pos_hits = sum(word in text for word in POSITIVE_WORDS) neg_hits = sum(word in text for word in NEGATIVE_WORDS) total = pos_hits + neg_hits if total == 0: return {"负面": 0.2, "中性": 0.6, "正面": 0.2} pos_prob = pos_hits / total neg_prob = neg_hits / total return { "负面": round(0.1 + 0.8 * neg_prob, 3), "中性": 0.1, "正面": round(0.1 + 0.8 * pos_prob, 3), } def analyze_sentiment(payload: SentimentRequest) -> SentimentResponse: prob_dict = _simple_score(payload.text) label = max(prob_dict.items(), key=lambda x: x[1])[0] return SentimentResponse( request_id=payload.request_id, probabilities=prob_dict, label=label, model_version="rule-0.1", ) def analyze_affective_state(payload: AffectiveAnalysisRequest) -> AffectiveAnalysisResponse: weights: Dict[str, float] = {} for sig in payload.affective_signals: weights[sig.emotion.lower()] = weights.get(sig.emotion.lower(), 0.0) + sig.intensity if not weights: dominant, confidence = "neutral", 0.2 else: dominant, score = max(weights.items(), key=lambda item: item[1]) total = sum(weights.values()) confidence = score / total if total else 0.0 if dominant in {"frustration", "anxious", "沮丧"}: message = f"检测到 {payload.student_id} 在「{payload.current_task}」中可能感到挫折。建议先处理最关键步骤。" elif dominant in {"bored", "disengaged"}: message = f"{payload.student_id} 的情绪趋于低唤醒，可尝试切换更具挑战性的子任务。" elif dominant in {"confident", "excited", "积极"}: message = f"{payload.student_id} 状态积极，可趁势加入举一反三的问题。" else: message = f"{payload.student_id} 情绪较为平稳，保持当前节奏并轻量检查理解情况。" if payload.recent_performance: message += f" 学习表现备注：{payload.recent_performance}。" database.log_emotion(payload.student_id, dominant, confidence, message) state = AffectiveState( dominant_emotion=dominant, confidence=round(confidence, 3), message=message, regulation_strategies=[ "使用 1-2 句同理心话语回应学生感受。", f"根据情绪状态对「{payload.current_task}」调整脚手架层级。", ], ) nudges = [ "给出分步提示或降低任务难度。", "安排 2 分钟休息或切换轻量任务。", ] return AffectiveAnalysisResponse( request_id=payload.request_id, student_id=payload.student_id, state=state, nudges=nudges, model_version="rule-0.1", )