基于改进Markov算法的人类活动信息挖掘

Human Activity Information Mining Based on the Improved Markov Algorithm

近年来,基于位置的社交媒体飞速发展,为人类移动规律的挖掘与研究带来新的数据源。基于扩展Markov模型,加入时间维度,提出一种利用社交媒体时空数据挖掘人类活动规律的方法,探索用户的活动位置和活动位置的变化规律。应用该方法对北京市新浪微博用户的个体和群体活动规律进行探索,可有效挖掘人类在以小时为单位细粒度时段的移动规律并由此反映区位人口的动态变化。

In recent years, the rapid development of location-based social media has provided large and longer-term human activity data, which brings a new data source for the excavation and research of human movement laws. In this paper, we extended the Markov model by joining the time dimension, and proposed a method for mining human activity patterns using social media spatio-temporal data, to explore the user's activity location and the change rule of the location. This method was used to explore the rules of individual and group activities of Sina Weibo users in Beijing. The experimental results show that this method can effectively extract the movement rule of human in hourly fine granularity period and reflect the dynamic change of regional population.