rerank-service

Lin-A1's avatarfrom Lin-A1

文档重排序服务,根据查询对候选文档进行相关性评分和排序

⭐1stars🔀0forks📁View on GitHub🕐Updated Jan 8, 2026⬇️3downloads

When & Why to Use This Skill

The rerank-service is a specialized document re-ranking tool designed to significantly enhance the precision of information retrieval systems. By employing advanced deep learning models (such as BGE-Reranker), it performs fine-grained semantic analysis to score and re-order candidate documents against a specific query. This ensures that the most relevant information is prioritized, effectively boosting Top-K accuracy for RAG (Retrieval-Augmented Generation) pipelines and complex search tasks.

Use Cases

  • RAG Pipeline Optimization: Refine initial retrieval results from vector databases to ensure the LLM receives the most contextually accurate information, reducing hallucinations.
  • Enterprise Search Enhancement: Improve the quality of internal knowledge base searches by re-ranking keyword-based or semantic search results for higher relevance.
  • Context Window Management: Select only the most pertinent document snippets to fit within an agent's context window, maximizing information density and reducing processing costs.
  • Information Filtering: Automatically filter out noise and irrelevant documents from large-scale data retrieval tasks in research and analysis workflows.
namererank-service
description文档重排序服务(Reranker)。基于深度学习模型对检索候选结果进行细粒度相关性打分与重新排序,显著提升检索结果的精准度(Top-K 准确率)。

功能

根据查询语句对候选文档进行相关性评分和排序,提升检索准确性。

调用方式

from services.rerank_service.client import RerankServiceClient

client = RerankServiceClient()

query = "什么是机器学习?"
documents = [
    "机器学习是人工智能的一个分支,通过数据训练模型。",
    "今天天气很好,适合出去散步。",
    "深度学习是机器学习的子领域,使用神经网络。"
]

# 完整重排序结果
result = client.rerank(query, documents, top_n=2)

# 简化结果:(索引, 分数, 文档) 元组列表
ranked = client.rerank_documents(query, documents, top_n=2)

# 只获取最相关的文档索引
indices = client.get_top_indices(query, documents, top_n=2)  # -> [0, 2]

返回格式

{
  "id": "rerank-xxx",
  "model": "BAAI/bge-reranker-v2-m3",
  "results": [
    {
      "index": 0,
      "document": {"text": "机器学习是人工智能的一个分支..."},
      "relevance_score": 0.999
    },
    {
      "index": 2,
      "document": {"text": "深度学习是机器学习的子领域..."},
      "relevance_score": 0.098
    }
  ]
}