#!/usr/bin/env python3 """ 媒体评估脚本 功能： 1. 智能采集媒体基本信息（历史、规模、影响力） 2. 评估媒体政治倾向 3. 生成媒体画像 """ import json import re from typing import Dict, Any, Optional, List from dataclasses import dataclass @dataclass class MediaProfile: """媒体画像""" name: str url: str established_year: Optional[int] headquarters: Optional[str] owner: Optional[str] political_leaning: str credibility_score: float # 0-100 audience_scale: str # "global", "regional", "national", "local" language: List[str] media_type: str # "news_agency", "newspaper", "tv", "online", "magazine" notes: List[str] # 国际主流媒体数据库（示例，实际应用中应使用外部知识库） INTERNATIONAL_MEDIA_DB = { "reuters.com": { "name": "路透社 (Reuters)", "established": 1851, "headquarters": "伦敦，英国", "owner": "汤森路透", "political_leaning": "中间偏中立", "credibility": 88, "audience_scale": "global", "media_type": "news_agency", "description": "世界最大的国际多媒体新闻通讯社之一，以快速、准确、中立著称。" }, "apnews.com": { "name": "美联社 (AP)", "established": 1846, "headquarters": "纽约，美国", "owner": "非营利合作组织", "political_leaning": "中立", "credibility": 90, "audience_scale": "global", "media_type": "news_agency", "description": "美国最大的非营利性新闻机构，以客观、事实驱动的报道闻名。" }, "bbc.com": { "name": "英国广播公司 (BBC)", "established": 1922, "headquarters": "伦敦，英国", "owner": "英国公共媒体", "political_leaning": "中间偏左（社会自由主义）", "credibility": 85, "audience_scale": "global", "media_type": "broadcast", "description": "英国公共广播机构，全球最具影响力的媒体之一。" }, "cnn.com": { "name": "美国有线电视新闻网 (CNN)", "established": 1980, "headquarters": "亚特兰大，美国", "owner": "华纳兄弟探索", "political_leaning": "中间偏左（自由主义）", "credibility": 72, "audience_scale": "global", "media_type": "tv", "description": "美国主要有线电视新闻网络，以24小时滚动新闻著称。" }, "nytimes.com": { "name": "纽约时报 (The New York Times)", "established": 1851, "headquarters": "纽约，美国", "owner": "纽约时报公司", "political_leaning": "自由主义/进步主义", "credibility": 82, "audience_scale": "global", "media_type": "newspaper", "description": "美国最具影响力的报纸之一，深度调查报道著称。" }, "foxnews.com": { "name": "福克斯新闻 (Fox News)", "established": 1996, "headquarters": "纽约，美国", "owner": "福克斯公司", "political_leaning": "保守主义", "credibility": 60, "audience_scale": "national", "media_type": "tv", "description": "美国主要有线电视新闻网络，以保守主义观点著称。" }, "theguardian.com": { "name": "卫报 (The Guardian)", "established": 1821, "headquarters": "伦敦，英国", "owner": "斯科特信托基金会", "political_leaning": "中间偏左（社会民主主义）", "credibility": 84, "audience_scale": "global", "media_type": "newspaper", "description": "英国主要报纸之一，以进步主义立场和深度调查著称。" }, "wsj.com": { "name": "华尔街日报 (The Wall Street Journal)", "established": 1889, "headquarters": "纽约，美国", "owner": "新闻集团", "political_leaning": "商业保守主义/中右翼", "credibility": 83, "audience_scale": "global", "media_type": "newspaper", "description": "美国主要金融报纸，以商业和金融新闻著称。" }, } def extract_domain(url: str) -> str: """从URL提取域名""" if not url: return "" # 移除协议前缀 url = re.sub(r'^https?://', '', url) # 移除www前缀 url = re.sub(r'^www\.', '', url) # 移除路径 url = url.split('/')[0] return url.lower() def assess_media_basic_info(media_url: str, media_name: str = None) -> Dict[str, Any]: """ 评估媒体基本信息 Args: media_url: 媒体URL media_name: 媒体名称（可选） Returns: Dict: 媒体基本信息 """ domain = extract_domain(media_url) # 查询数据库 if domain in INTERNATIONAL_MEDIA_DB: db_info = INTERNATIONAL_MEDIA_DB[domain] return { "source": "database", "media_name": db_info["name"], "established_year": db_info["established"], "headquarters": db_info["headquarters"], "owner": db_info["owner"], "audience_scale": db_info["audience_scale"], "media_type": db_info["media_type"], "description": db_info["description"], } # 如果不在数据库中，返回基本信息 return { "source": "inferred", "media_name": media_name or "未知媒体", "established_year": None, "headquarters": None, "owner": None, "audience_scale": "unknown", "media_type": "unknown", "description": f"媒体 URL: {media_url}", } def assess_political_leaning(media_url: str, media_name: str = None) -> Dict[str, Any]: """ 评估媒体政治倾向 Returns: Dict: 政治倾向评估 - leaning: 政治倾向标签 - score: 评分（0-100，左翼=0，右翼=100，中立=50） - evidence: 评估依据 - confidence: 置信度 """ domain = extract_domain(media_url) # 查询数据库 if domain in INTERNATIONAL_MEDIA_DB: leaning_text = INTERNATIONAL_MEDIA_DB[domain]["political_leaning"] # 将文本倾向转换为评分 score_map = { "中立": 50, "中间偏中立": 50, "自由主义/进步主义": 25, "中间偏左（社会自由主义）": 40, "中间偏左（社会民主主义）": 35, "保守主义": 75, "商业保守主义/中右翼": 65, } score = score_map.get(leaning_text, 50) return { "leaning": leaning_text, "score": score, "evidence": f"基于国际媒体数据库对 {INTERNATIONAL_MEDIA_DB[domain]['name']} 的历史记录", "confidence": "high", "source": "database", } # 默认评估 return { "leaning": "未知", "score": 50, "evidence": "媒体不在已知数据库中，无法确定政治倾向", "confidence": "low", "source": "inferred", } def assess_media_credibility(media_url: str) -> Dict[str, Any]: """ 评估媒体可信度 Returns: Dict: 可信度评估 - score: 0-100分 - level: "high", "medium", "low" - factors: 评估因素 """ domain = extract_domain(media_url) if domain in INTERNATIONAL_MEDIA_DB: credibility_score = INTERNATIONAL_MEDIA_DB[domain].get("credibility", 60) else: # 未知媒体给予中等分数 credibility_score = 60 if credibility_score >= 80: level = "high" elif credibility_score >= 60: level = "medium" else: level = "low" return { "score": credibility_score, "level": level, "factors": { "historical_accuracy": credibility_score if domain in INTERNATIONAL_MEDIA_DB else 60, "editorial_standards": credibility_score if domain in INTERNATIONAL_MEDIA_DB else 60, "corrections_policy": credibility_score if domain in INTERNATIONAL_MEDIA_DB else 60, } } def generate_media_assessment(media_url: str, article_url: str = None, media_name: str = None) -> Dict[str, Any]: """ 生成完整的媒体评估 Args: media_url: 媒体主页URL article_url: 文章URL（可选，用于进一步分析） media_name: 媒体名称（可选） Returns: Dict: 完整的媒体评估结果 - basic_info: 基本信息 - political_leaning: 政治倾向 - credibility: 可信度 - summary: 总结性描述 """ basic_info = assess_media_basic_info(media_url, media_name) political_leaning = assess_political_leaning(media_url, media_name) credibility = assess_media_credibility(media_url) # 生成总结 media_display_name = basic_info.get("media_name", media_name or "未知媒体") summary = f"{media_display_name}" if basic_info.get("established_year"): summary += f"，创立于{basic_info['established_year']}年" if basic_info.get("headquarters"): summary += f"，总部位于{basic_info['headquarters']}" summary += f"。政治倾向：{political_leaning['leaning']}，可信度评分：{credibility['score']}/100" return { "media_url": media_url, "article_url": article_url, "basic_info": basic_info, "political_leaning": political_leaning, "credibility": credibility, "summary": summary, "generated_at": "2024-04-02T00:00:00Z" # 实际应用中应使用当前时间 } def main(): """命令行入口，用于测试""" import sys # 示例 if len(sys.argv) > 1: media_url = sys.argv[1] article_url = sys.argv[2] if len(sys.argv) > 2 else None media_name = sys.argv[3] if len(sys.argv) > 3 else None else: media_url = "https://www.reuters.com" article_url = None media_name = "路透社" result = generate_media_assessment(media_url, article_url, media_name) print(json.dumps(result, ensure_ascii=False, indent=2)) if __name__ == "__main__": main()