基于用户评论的农业新技术推荐模型设计与应用展望  

Design and Application Prospect of Agricultural New Technology Recommendation Model Based on User Reviews

作　　者：刘伟[1] 刘世洪[1] 王翠 宋林鹏 Liu Wei;Liu Shihong;Wang Cui;Song Linpeng(Agricultural Information Institute,Chinese A cademy of Agricultural Sciences,Key Laboratory of Agri-information Service Technology,Ministry of Agriculture and Rural Affairs,Beijing 100081)

机构地区：[1]中国农业科学院农业信息研究所/农业农村部信息服务技术重点实验室,北京100081

出　　处：《农业展望》2023年第8期100-105,共6页Agricultural Outlook

摘　　要：近年来,推荐系统的应用取得了飞速进步。大数据、人工智能技术的出现为农业信息化的加速发展提供了广阔的空间和前景。为提升农业领域内推荐技术的应用,满足农业用户的信息获取需求,对传统协同过滤推荐算法进行了一定的改进,重点体现在融合了K-means算法以及BIRCH算法进行聚类分析,通过搭建HowNet极性词典解决传统协同过滤方法过度依赖用户具体评分的问题,并提出了一种个性化推荐模型,利用相关数据源,进行模型验证。实验结果表明,该模型运行稳定,可以达到精准推荐农业技术信息的目的。In recent years, the application of recommendation system has made rapid progress. The emergence of big data and artificial intelligence technology provides a broad space and prospect for the accelerated development of agricultural informatization. In order to improve the application of recommendation technology in the agricultural field and meet the information acquisition needs of agricultural users, the traditional collaborative filtering recommendation algorithm was improved to some extent, focusing on the integration of K-means algorithm and BIRCH algorithm for cluster analysis, through the construction of HowNet polarity dictionary to solve the problem that traditional collaborative filtering methods over-rely on users' specific ratings, and a personalized recommendation model was proposed,using related data sources to verify the model. The experimental results showed that the model runs stably and can achieve the purpose of accurately recommending agricultural technical information.

关 键 词：协同过滤算法 聚类分析 HowNet极性词典 个性化推荐模型 

分 类 号：S126[农业科学—农业基础科学] TP391.3[自动化与计算机技术—计算机应用技术]