Resume Parser

匹配度评分严格规则

1. 90-100分：完全匹配：所有核心要求100%满足，加分要求满足80%以上，有超出要求的亮点
2. 70-89分：基本匹配：核心要求全部满足，加分要求满足50%以上，无明显核心短板
3. 60-69分：勉强匹配：核心要求基本满足，有1项非关键核心要求不满足，加分要求满足30%以上，可以进入面试
4. <60分：不匹配：核心要求有2项及以上不满足，或有1项关键核心要求不满足，不符合岗位基本要求

核心要求判定规则

• 岗位JD中明确标注「必须」「要求」「需具备」的技能/经验
• 岗位名称对应的核心能力（如AI算法岗必须懂深度学习，后端岗必须懂编程语言）
• 明确的工作年限、学历要求

脚本使用

1. PDF文本提取

python scripts/extract_pdf.py <input-pdf-path>

返回纯文本内容

2. Word文本提取

python scripts/extract_docx.py <input-docx-path>

返回纯文本内容

3. 图片OCR提取

python scripts/extract_image.py <input-image-path>

返回OCR识别的文本内容

4. 结构化解析

python scripts/parse_resume.py <extracted-text-file>

返回结构化JSON数据

5. 匹配度分析

python scripts/match_jd.py <resume-json-path> <jd-text-path>

返回匹配度分析结果

输出格式规范

结构化简历JSON格式

{
  "basic_info": {
    "name": "",
    "phone": "",
    "email": "",
    "age": null,
    "gender": "",
    "location": "",
    "work_years": null
  },
  "education": [
    {
      "school": "",
      "major": "",
      "degree": "",
      "start_date": "",
      "end_date": "",
      "gpa": "",
      "courses": []
    }
  ],
  "work_experience": [
    {
      "company": "",
      "position": "",
      "start_date": "",
      "end_date": "",
      "description": "",
      "achievements": [],
      "technologies": []
    }
  ],
  "projects": [
    {
      "name": "",
      "role": "",
      "start_date": "",
      "end_date": "",
      "description": "",
      "technologies": [],
      "achievements": []
    }
  ],
  "skills": {
    "technical": [],
    "soft": [],
    "languages": []
  },
  "certificates": [],
  "awards": [],
  "self_assessment": ""
}

匹配度分析格式

{
  "overall_score": 0-100,
  "dimensions": [
    {
      "name": "核心技能匹配",
      "score": 0-100,
      "weight": 0.4,
      "matched": ["匹配的核心技能列表"],
      "missing": ["缺失的核心技能列表"],
      "analysis": "详细分析说明"
    },
    {
      "name": "岗位职责匹配",
      "score": 0-100,
      "weight": 0.3,
      "matched": ["匹配的职责经验"],
      "gap": "职责差距描述",
      "analysis": "详细分析说明"
    },
    {
      "name": "经验/资历匹配",
      "score": 0-100,
      "weight": 0.15,
      "matched": ["匹配的经验点"],
      "gap": "经验差距描述",
      "analysis": "详细分析说明"
    },
    {
      "name": "学历/背景匹配",
      "score": 0-100,
      "weight": 0.15,
      "matched": "匹配结果描述",
      "gap": "背景差距描述（如果有）",
      "analysis": "详细分析说明"
    }
  ],
  "overall_analysis": "整体匹配情况总结，明确说明是否匹配",
  "strengths": ["简历优势列表"],
  "weaknesses": ["简历不足列表，必须明确核心差距"],
  "suggestions": ["具体的优化建议列表"]
}

依赖安装

首次使用前安装依赖：

pip install PyPDF2 python-docx pytesseract pillow python-multipart

注意：OCR功能需要安装Tesseract引擎