Abstract: This paper, from the unique perspective of predictable versus unpredictable announcement timing, examines investor information behaviors in a data-driven context. Using online search indices of listed companies and user behavior data from securities service apps, it empirically constructs and analyzes a novel measure of investor attention. The results show that for financial statement-related announcements, unpredictable announcement timing generates greater investor attention compared to predictable timing. Moreover, when announcement timing is unpredictable, financial statement-related announcements attract more investor attention than non-financial statement announcements. Furthermore, under unpredictable timing, different types of announcements exhibit a complementary effect, jointly influencing investor attention behavior.

Key words: data-driven, investor attention, information behavior, complementary effect

摘要： 本文从公告时间可预期与不可预期的视角出发，系统考察数据驱动背景下投资者的信息行为特征。研究基于上市公司网络搜索指数及证券服务类 APP的用户行为数据，构建新型投资者关注度量指标并进行实证检验。结果表明：对于报表类公告,发布时间不可预期时比可预期时更能引发投资者关注；在公告发布时间不可预期的情况下，报表类公告较非报表类公告更易吸引投资者关注；此外，当公告发布时间不可预期时，不同类型公告之间存在互补效应，共同影响投资者的关注行为。

关键词: 数据驱动, 投资者关注, 信息行为, 互补效应