基于向量空間模型的科技情報用戶畫像及場景化服務(wù)推送研究

王益成 王萍 張禹

摘 要：[目的]為了滿足情報3.0時代用戶對科技情報服務(wù)個性化與智能化的要求，采用用戶畫像方法挖掘科技情報用戶情報需求與偏好。[方法]以TAM理論的兩個關(guān)鍵因素（感知有用性與易用性）為理論基礎(chǔ)，通過挖掘科技情報用戶行為日志數(shù)據(jù)以及問卷調(diào)查等方法深入挖掘科技情報用戶在不同情境下的科技情報需求偏好以及搜索行為習(xí)慣，以專家訪談法為基礎(chǔ)，基于VSM理論構(gòu)建科技情報用戶畫像模型。[結(jié)果]通過協(xié)同過濾算法對不同場景的科技情報用戶進(jìn)行服務(wù)推送。[結(jié)論]運用標(biāo)簽云技術(shù)（Tagxedo技術(shù)）為處于不同場景的科技情報用戶定制有針對性的個性化服務(wù)方案是情報3.0時代一種有效的智能化服務(wù)途徑。

關(guān)鍵詞：情報3.0;用戶畫像;場景化推送;個性化服務(wù);向量空間模型;科技情報服務(wù)

DOI：10.3969/j.issn.1008-0821.2020.02.001

〔中圖分類號〕G252.0 〔文獻(xiàn)標(biāo)識碼〕A 〔文章編號〕1008-0821（2020）02-0003-08

Research on User Portrait and Scenario Recommendation of

Sci-tech Information Service Based on VSM

Wang Yicheng1 Wang Ping1 Zhang Yu2

（1.School of Management，Jilin University，Changchun 130022，China;

2.Department of Education，Northeast Normal University，Changchun 130024，China）

Abstract：[Purpose]In order to meet the requirements of for personalized and intelligent service of scientific and technological information in the period of intelligence 3.0，the user portrait method was adopted to explore the sci-tec users needs and preferences of scientific and technological information.[Method]Based on two key factors of TAM theory perceived usefulness and ease of use，through digging science and technology intelligence，users log data and questionnaire survey.The paper deeply explored the science and technology information users science and technology intelligence demand preference and search behavior in different situation，based on the expert interviews，the paper built science and technology intelligence user portrait model based on VSM theory.[Results]Collaborative Filtering Algorithm was used to recommend services to scientific and technological intelligence users in different scenarios.[Conclusion]Using tag cloud technology（tagxedo technology）to customize targeted personalized service solutions for scientific and technological users in different scenarios was an effective intelligent service channel in the intelligence 3.0 era.

Key words：intelligence 3.0;user portrait;scenario recommendation;personalized services;vector space model;sci-tec intelligence services