feat: 关键词助手的流程实现

server.py

@@ -0,0 +1,53 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+# @Time    : 2024/6/5 上午7:24
+# @Author  : Tim-Saijun https://zair.top
+# @File    : langchain_tutorials.py
+# @Project : SAgent
+
+from typing import List
+
+from fastapi import FastAPI
+from langchain_core.prompts import ChatPromptTemplate
+from langchain_core.output_parsers import StrOutputParser
+from langchain_openai import ChatOpenAI
+from langserve import add_routes
+
+# 1. Create prompt template
+system_template = "Translate the following into {language}:"
+prompt_template = ChatPromptTemplate.from_messages([
+    ('system', system_template),
+    ('user', '{text}')
+])
+
+# 2. Create model
+model = ChatOpenAI()
+
+# 3. Create parser
+parser = StrOutputParser()
+
+# 4. Create chain
+chain = prompt_template | model | parser
+
+
+# 4. App definition
+app = FastAPI(
+  title="LangChain Server",
+  version="1.0",
+  description="A simple API server using LangChain's Runnable interfaces",
+)
+
+# 5. Adding chain route
+
+add_routes(
+    app,
+    chain,
+    path="/chain",
+)
+
+if __name__ == "__main__":
+    import uvicorn
+
+    uvicorn.run(app, host="localhost", port=8000)
+
+

关键词助手.ipynb

@@ -0,0 +1,238 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "id": "initial_id",
+   "metadata": {
+    "collapsed": true,
+    "ExecuteTime": {
+     "end_time": "2024-06-05T09:41:38.526785Z",
+     "start_time": "2024-06-05T09:41:37.319136Z"
+    }
+   },
+   "source": [
+    "from IPython.display import display, Markdown\n",
+    "\n",
+    "from langchain_core.prompts import ChatPromptTemplate\n",
+    "from langchain_core.output_parsers import StrOutputParser\n",
+    "from langchain_openai import ChatOpenAI\n",
+    "\n",
+    "model = ChatOpenAI(model = \"gpt-4o\",\n",
+    "                   temperature = 0.1,\n",
+    "                   max_tokens = 4096,\n",
+    "                   # presence_penalty = 1,\n",
+    "                   )\n",
+    "parser = StrOutputParser()"
+   ],
+   "outputs": [],
+   "execution_count": 83
+  },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "# 1. 关键词筛选",
+   "id": "30ff1e47cd72de9f"
+  },
+  {
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-05T09:51:52.767269Z",
+     "start_time": "2024-06-05T09:51:52.764220Z"
+    }
+   },
+   "cell_type": "code",
+   "source": [
+    "# 1. Create prompt template\n",
+    "filter_system_template = \"\"\"\n",
+    "进行关键词分析，筛选出拓展词列表中与目标关键词高度相关的关键词，具体要求如下：\n",
+    "- 根据目标关键词，评估拓展词列表中每个关键词的相关性，删除相关性不高的关键词。\n",
+    "- 删除拓展词列表中的所有品牌词。\n",
+    "- 删除拓展词列表中的所有平台名称。\n",
+    "- 搜索意图匹配：确保每个关键词都与目标用户的搜索意图相匹配。不符合目标用户搜索意图的关键词应予以删除。\n",
+    "- 长尾关键词优先：倾向于保留具有具体意图且通常竞争较小的长尾关键词。\n",
+    "- 内容覆盖度：确保所选关键词能够被现有内容或计划中的内容充分覆盖，与网站内容策略相符合。\n",
+    "## Output:\n",
+    "直接输出筛选后的关键词列表，每行输出一个关键词。\n",
+    "\"\"\"\n",
+    "filter_prompt_template = ChatPromptTemplate.from_messages([\n",
+    "    ('system', filter_system_template),\n",
+    "    ('user', '目标关键词：{corekeywords}  拓展词列表：{relatedkeywords}')\n",
+    "])\n"
+   ],
+   "id": "7059fa2d895f586d",
+   "outputs": [],
+   "execution_count": 141
+  },
+  {
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-05T09:50:55.619759Z",
+     "start_time": "2024-06-05T09:50:55.613906Z"
+    }
+   },
+   "cell_type": "code",
+   "source": [
+    "filter_prompt_template.invoke({'corekeywords': \"打印机\",\n",
+    "              'relatedkeywords': \"激光打印机 喷墨打印机\"})"
+   ],
+   "id": "98fd18304ff75856",
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "ChatPromptValue(messages=[SystemMessage(content='\\n进行关键词分析，筛选出拓展词列表中与目标关键词高度相关的关键词，具体要求如下：\\n- 根据目标关键词，评估拓展词列表中每个关键词的相关性，删除相关性不高的关键词。\\n- 删除拓展词列表中的所有品牌词。\\n- 删除拓展词列表中的所有平台名称。\\n- 搜索意图匹配：确保每个关键词都与目标用户的搜索意图相匹配。不符合目标用户搜索意图的关键词应予以删除。\\n- 长尾关键词优先：倾向于保留具有具体意图且通常竞争较小的长尾关键词。\\n- 内容覆盖度：确保所选关键词能够被现有内容或计划中的内容充分覆盖，与网站内容策略相符合。\\n## Output:\\n直接输出筛选后的关键词列表，每行输出一个关键词。\\n'), HumanMessage(content='目标关键词：打印机  拓展词列表：激光打印机 喷墨打印机')])"
+      ]
+     },
+     "execution_count": 118,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "execution_count": 118
+  },
+  {
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-05T09:53:52.219922Z",
+     "start_time": "2024-06-05T09:53:51.496246Z"
+    }
+   },
+   "cell_type": "code",
+   "source": [
+    "chain = filter_prompt_template | model | parser\n",
+    "filter_res = chain.invoke({'corekeywords': \"打印机\",\n",
+    "              'relatedkeywords': \"激光打印机 喷墨打印机 汉堡王 京东商城 亚马逊 活字印刷术 印刷制品\"})\n",
+    "display(Markdown(filter_res))"
+   ],
+   "id": "25cc895e34e2ac5e",
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<IPython.core.display.Markdown object>"
+      ],
+      "text/markdown": "激光打印机  \n喷墨打印机"
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "execution_count": 149
+  },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "# 2. 筛选结果确认",
+   "id": "2569efb1c53bffd9"
+  },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "# 3. 关键词分类",
+   "id": "792a4d9ab769d833"
+  },
+  {
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-05T08:41:57.856976Z",
+     "start_time": "2024-06-05T08:41:57.851765Z"
+    }
+   },
+   "cell_type": "code",
+   "source": [
+    "classify_system_template = \"\"\"\n",
+    "## Role:\n",
+    "你是一名具有二十年工作经验的Google SEO专家，擅长快速制定SEO策略。\n",
+    "## Background:\n",
+    "我需要快速进行关键词分类，请参考用户提供的关键词列表，并按照指定的维度进行分类。\n",
+    "## Constraints:\n",
+    "1. 关键词列表: {key_words}\n",
+    "\n",
+    "2. 关键词分类维度: 颜色、规格、属性、材质、功能、样式、型号、品牌、商业词、地域词、文章词。\n",
+    "3. 以下关键词均属于商业词类别: manufacture, supplier, wholesale, OEM, ODM, company, factory, private label, Custom, bulk, Import, Export, Agent, Dealer, distributor。\n",
+    "## Workflows：\n",
+    "1. 将关键词列表中的每个词按照[Constraints]中关键词分类维度进行分类。\n",
+    "2. 单个关键词只能隶属于一个关键词分类维度，不能同时属于多个分类。\n",
+    "3. 每个关键词必须进行一次分类，不得遗漏。\n",
+    "4. 分类完毕后，在对应的分类字段中填写关键词中的修饰词。\n",
+    "5. 确保关键词列表中的每个词都被执行分类。\n",
+    "## Output:\n",
+    "1. 请用表格输出。\n",
+    "2. 不要任何解释，直接输出表格。\n",
+    "## Output Examples:\n",
+    "| 目标关键词 | 颜色 | 规格 | 属性 | 材质 | 功能 | 样式 | 型号 | 品牌 | 商业词 | 地域词 | 文章词 |\n",
+    "|------------|------|------|------|------|------|------|------|------|--------|--------|--------|\n",
+    "| 关键词1    |      |      |      |      |      |      |      |      |        |        |        |\n",
+    "| 关键词2    |      |      |      |      |      |      |      |      |        |        |        |\n",
+    "| 关键词3    |      |      |      |      |      |      |      |      |        |        |        |\n",
+    "## Notice:\n",
+    "你必须将关键词列表中的每个关键词全部执行一次关键词分类，不得遗漏。每个关键词只能属于一个分类维度，不能同时出现在多个分类中。在分类字段中只填写关键词中的修饰词。\n",
+    "## Initialization:\n",
+    "作为[Role]，在[Background]下，回顾你的技能，严格遵守[Constraints]进行关键词分类，按[Workflows]执行流程，按[Output]格式要求输出。\n",
+    "## Additional Constraints:\n",
+    "1. 确保所有关键词都被分类，不得遗漏。\n",
+    "2. 分类完毕后，检查表格，确保每个关键词都被准确分类。\n",
+    "3. 用户输入的目标关键词中的任何单词不参与关键词分类。\n",
+    "\"\"\"\n",
+    "classify_prompt_template = ChatPromptTemplate.from_messages([\n",
+    "    ('system', classify_system_template),\n",
+    "    ('user', '目标关键词：{target_keywords}')\n",
+    "])\n",
+    "# res = classify_prompt_template.invoke({'target_keywords': \"打印机\", 'key_words': \"激光打印机 喷墨打印机\"})\n",
+    "# res.to_messages()"
+   ],
+   "id": "2c72b992f3a15f94",
+   "outputs": [],
+   "execution_count": 40
+  },
+  {
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-05T09:54:05.635944Z",
+     "start_time": "2024-06-05T09:54:03.158609Z"
+    }
+   },
+   "cell_type": "code",
+   "source": [
+    "classify_chain = classify_prompt_template | model | parser\n",
+    "class_res = classify_chain.invoke({'target_keywords': \"打印机\", 'key_words': filter_res})\n",
+    "display(Markdown(class_res))"
+   ],
+   "id": "2db5d8e8d991213e",
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<IPython.core.display.Markdown object>"
+      ],
+      "text/markdown": "| 目标关键词 | 颜色 | 规格 | 属性 | 材质 | 功能 | 样式 | 型号 | 品牌 | 商业词 | 地域词 | 文章词 |\n|------------|------|------|------|------|------|------|------|------|--------|--------|--------|\n| 激光打印机 |      |      |      |      | 激光 |      |      |      |        |        |        |\n| 喷墨打印机 |      |      |      |      | 喷墨 |      |      |      |        |        |        |"
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "execution_count": 150
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 2
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython2",
+   "version": "2.7.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}