摘要
[目的/意义]地理标志是中国地方政府开展品牌建设的重要抓手,政府在地理标志农产品品牌管理领域的政策注意力问题值得深入探究。论文旨在构建地理标志品牌分析模型,以支持分析品牌发展现状、探究政策注意力分布、提出优化策略。[方法/过程]论文首先采集数据构建语料库,采用UIE通用信息抽取框架抽取文本,采用基于Transformer模型的品牌管理措施分类模型,开展管理措施分类,最终完成构建品牌分析模型,支持政策注意力分析,找到瓶颈问题,进行措施优化等应用。[结果/结论]论文以烟台苹果为例进行模型验证,对大量语料进行抽取、分类,发现烟台苹果政策注意力较集中、措施雷同性高,收集文本中的41.1%集中在品牌定位与规划部分,31.7%集中在品牌核心能力提升部分,涉及品牌营销及保护的不到10%,未利用好地理标志品牌的品牌效应。
[Purpose/Significance]Geographical indications(GIs)are an important tool for local governments in China to carry out brand building of agricultural products.Brand management is a continuous systematic project involving multiple subjects.Among them,the problem of government policy attention in the field of brand management of agricultural products with GIs deserves in-depth study.This paper aims to construct a policy attention analysis model in the field of brand management of GI agricultural products based on natural language processing technology.This model provides technical support for local governments to explore the status quo of local GI agricultural products'brand management,analyze the distribution of policy attention,and assist in optimizing strategies to promote their products'brand development.[Method/Process]The study is focused on the distribution of attention paths of GI agricultural products'brand management policies from the perspective of the local government:an analysis model of brand management policies of GI agricultural products was constructed in order to support the local government to carry out the analysis of the status quo of local agricultural products'brand management and policy optimization,and provide decision-making support for the optimization of brand management measures of the local agricultural products.First,this paper built a basic corpus based on the Python crawler technology,collected authoritative public information on the Internet,utilized the domain dictionary and UIE general information extraction framework to extract the text of management measures published in local government policies,and built a database of brand management measures of GI agricultural product.Second,a classification model of brand management measures of GI agricultural products based on the Transformer model was constructed.Third,this paper built a classification model based on the Transformer model,which can classify the extracted brand management measures of agricultural products and construct the policy attention distribution map.Finally,based on the policy attention distri-bution given by the model,we can find the brand management bottlenecks and recommend countermeasures to solve the bottlenecks.[Results/Conclusions]This paper takes Yantai apples as an example for model validation.After extracting and categorizing Yantai apples'brand management data,it is found that the policy attention of Yantai apples is more con-centrated,and the measures are highly similar,with 41.1%of the text of the measures focused on the part of brand positioning and planning,31.7%on the part of brand competitiveness enhancement,and less than 10%on the part of brand marketing and protection.It can be seen that the brand effect of Yantai apples with GIs has not been well utilized.
作者
霍梦佳
刘娟
黄杰
HUO Mengjia;LIU Juan;Huang Jie(Institute of Agricultural Information,Chinese Academy of Agricultural Sciences,Beijing 100080;Key Laboratory of Agricultural Big Data,Ministry of Agriculture,Beijing 100080)
出处
《农业图书情报学报》
2023年第7期94-104,共11页
Journal of Library and Information Science in Agriculture
基金
中国农业科学院基本科研业务费项目“基于文本挖掘的地理标志农产品品牌管理模式识别技术研究与应用”(JBYW-AII-2022-33
Y2023XK13)
现代农业产业技术体系北京市创新团队建设项目“现代农业产业技术体系北京市创新团队建设”(BAIC10-2023-E10)。
关键词
农产品地理标志
自然语言处理
政策注意力
信息抽取
文本分类
geographical indications of agricultural products
natural language processing
policy attention
information extraction
text classification