摘要
为提高用户对外卖菜品的点击率与购买率,本研究设计了一款基于文本挖掘的智能推荐型外卖软件。该软件通过网络爬虫技术抓取用户在外卖平台上的购买信息,包括订单详情、用户评价等文本数据,然后利用自然语言处理技术对这些文本数据进行预处理,包括分词、去停用词等,以提取出有价值的文本特征,再通过特征选择算法从这些文本特征中筛选出最具代表性的特征,从而简化模型复杂度,提高推荐准确性。本研究采用了朴素贝叶斯分类器对处理后的文本特征进行文本挖掘,通过计算给定特征下不同类别的概率,朴素贝叶斯分类器可以预测用户的偏好类型,进而利用协同过滤算法实现个性化外卖菜品推荐。经实验验证可知,经该软件推荐的外卖菜品具有较高的购买率与点击率,同时该软件在推荐过程中可保持良好的归一化折损累计增益,能有效为用户推荐理想的外卖内容。
In order to improve the click rate and purchase rate of users’takeout dishes,the study designs an intelligent recommendation-based takeout software based on text mining.The software captures users’purchase information on the takeout platform through web crawler technology,including order details,user evaluation and other text data,and then utilizes natural language processing technology to pre-process these text data,including word splitting,de-duplication,etc.,in order to extract valuable text features,and then filters out the most representative features from these text features through a feature selection algorithm to simplify model complexity and improve recommendation accuracy.In this study,a plain Bayesian classifier is used to perform text mining on the processed text features.By calculating the probability of different categories under the given features,the plain Bayesian classifier can predict the type of user preferences,and then use collaborative filtering algorithms to achieve personalized takeout dish recommendation.After experimental verification,it can be seen that the takeaway dishes recommended by the software have a high purchase rate and click rate,and at the same time,good normalized discount cumulative gain can be maintained in the recommendation process,which can effectively recommend ideal takeaway contents for users.
作者
王峰
WANG Feng(Harbin Fangtian Network Technology Co.,Ltd.,Harbin Heilongjiang 150000,China)
出处
《信息与电脑》
2024年第16期162-165,共4页
Information & Computer
关键词
文本挖掘
智能推荐型
外卖软件设计
text mining
intelligent recommendation
takeaway software design
