摘要
在地方图书馆文献资源阅读推广过程中,无法处理稀疏性强的高维数据,使得最终推广结果的F1值较低。因此,构建地方图书馆文献资源阅读推广平台。首先,选择网络最大影响节点作为网络推广的核心,计算出地方图书馆文献资源热度。其次,集成用户行为数据构建用户画像,提取用户画像特征词。最后,计算用户画像特征词变量与文献资源变量之间的最大互信息系数,得出平台文献资源阅读个性化推广结果。测试结果表明,应用该平台后,推荐结果的F1值为0.8,基本满足文献资源个性化推广要求。
In the process of promoting the reading of literature resources in local libraries,it is difficult to handle highdimensional data with strong sparsity,resulting in a lower F1 value for the final promotion result.Therefore,building a local library literature resource reading and promotion platform.Firstly,select the node with the greatest impact on the network as the core of network promotion,and calculate the popularity of local library literature resources.Secondly,integrating user behavior data to construct user profiles and extracting feature words from user profiles.Finally,calculate the maximum mutual information coefficient between the variable of user profile feature words and the variable of literature resources,and obtain the personalized promotion results of platform literature resource reading.The test results show that after applying the platform,the F1 value of the recommendation result is 0.8,which basically meets the requirements of personalized promotion of literature resources.
作者
李爽
LI Shuang(Gansu Library,Lanzhou Gansu 730030,China)
出处
《信息与电脑》
2023年第22期85-87,共3页
Information & Computer
关键词
网络环境
地方图书馆
文献资源
阅读推广平台
network environment
local libraries
literature resources
reading promotion platform