【摘要】 针对当下垃圾分类的热门话题，从无到有搭建一个以垃圾分类为中心的一定规模的垃圾分类知识图谱。项目采用网络爬虫、BiLSTM-CRF、BiGRU-Attention等技术对垃圾分类中文文本数据进行爬虫获取、实体识别、关系抽取，形成“实体-关系-实体”的三元组信息，使用Neo4j可视化展示三元组关系，形成结构化的知识图谱。构建垃圾分类知识图谱，为垃圾分类知识问答系统奠定了基础，使人们在日常生活中能够更好地将垃圾分好类、分对类。更多还原

【Abstract】 In response to the current hot topic of garbage classification, a knowledge graph of garbage classification with a certain size centered on garbage classification has been built from scratch. The project used web crawlers, BiLSTM-CRF, BiGRU-Attention and other technologies to perform crawler acquisition, entity recognition, and relationship extraction on garbage classification Chinese text data to form "entity-relationship-entity" triples. It used Neo4 j to visualize triplet relationship, forming a structured knowledge graph. Construction of a garbage classification knowledge map laid the foundation for a garbage classification knowledge question answering system, so that people could better classify garbage into two categories in daily life.更多还原