摘要
[目的/意义]开展在线医疗社区用户的需求信息的聚类研究有利于更高效地提供信息推荐服务,对其信息优化以及提高搜索效率具有实际意义。[方法/过程]根据丁香园论坛近23个季度的热点需求词构造词频矩阵,将Logistic增长模型与词频变化率模型相结合对需求词进行二维分选。[结果/结论]通过逻辑曲线拟合以及词频变化率Z值的计算将丁香园论坛热点需求词划分为上升型的长期需求词、下降型的长期需求词、上升型的短期需求词以及下降型的短期需求词,为在线医疗社区提供更高效的信息推荐服务提供了新思路。
[Purpose/significance]The clustering research on demand information of online health community users is beneficial to provide information recommendation service more efficiently,and has practical significance for information optimization and improving search efficiency.[Method/process]This research constructs word frequency matrix according to the hot demand words of DXY Forum in the last 23 quarters,and sorts out these words in two dimensions combining the Logistic growth model with the word frequency change rate model.[Result/conclusion]Through logical curve fitting and calculation of word frequency change rate Z value,hot demand words in DXY Forum are divided into rising long-term demand words,declining long-term demand words,rising short-term demand words,and declining short-term demand words,which can provide new insights on providing more effective information recommendation services for online health communities.
作者
杨少梅
吴皓月
李胜利
Yang Shaomei;Wu Haoyue;Li Shengli(Department of Economic Management,North China Electric Power University,Baoding Hebei 071003;Department of Information Management,Peking University,Beijing 100871)
出处
《情报探索》
2022年第2期43-52,共10页
Information Research
