期刊文献+

智能检索环境下语义分词调整策略的研究 被引量:2

Research on Semantic Tokenization Adjustment Strategy in the Intelligent Searching Environment
下载PDF
导出
摘要 检索系统是专利审查检索重要的依托,优秀的检索系统有助于检索质量及效率的提升。国家知识产权局上线的智能检索环境集合了原有S系统和语义智能检索的优势,其中检索工具的核心为语义检索和语义排序,而这一步的关键点在于分词。本研究结合具体的案例,探讨了在智能检索环境中语义分词调整的策略,结论为在检索过程中一般不需要删除分词,可以增加部分关键组分的分词,而且相应分词的权重不宜过高,建议调整为2或3即可。 Searching system is an important basis for patent examination search, good searching system is helpful to improve the quality and efficiency of searching. The intelligent searching environment of the CNIPA has the advantage of the original S-system and semantic intelligent searching, while the semantic searching and semantic sorting is the core, and one of the key point lies in the tokenization. For this point, the adjustment strategy of semantic tokenization in the intelligent searching system was preliminarily discussed in this study combining with the concrete case, the conclusion is that generally the semantic tokenization do not need to be deleted in the searching process, part of semantic tokenization of the key components could be added, and the weight coefficient of the corresponding tokenization should not be too high, 2 or 3 is recommended.
作者 尹梦岩 王梦霞 YIN Mengyan;WANG Mengxia(Patent Examination Cooperation Guangdong Center of the Patent Office,CNIPA,Guangzhou 510000,China)
出处 《河南科技》 2022年第20期155-158,共4页 Henan Science and Technology
关键词 专利检索 智能检索环境 语义分词 权重 patent search intelligent searching environment semantic tokenization weight coefficient
  • 引文网络
  • 相关文献

参考文献1

共引文献5

同被引文献17

引证文献2

;
使用帮助 返回顶部