Learning trendiness from twitter to web： a comparative analysis of microblog and web trending topics

The development of microblog services has a considerable etfect on the patterns oI wed access and Internet resources discovery. Understanding the interrelation between information diffusion in online social media and user web interests can help the web ecosystem stakeholders in developing new services and designing efficient systems with optimized resources. This paper explores whether or not one can infer the trends of topics in the web by observing the Twitter microcosm. Using data- sets collected from Twitter and two representative web services （Google and Alexa）, this work con- ducts a comparative analysis between trending patterns of topics in Twitter and in the web by consid- ering both the temporal and spatial perspectives, and finds that individual topics in Twitter and in the web share similar trending patterns both from the temporal and spatial aspects. Nevertheless, the tren- diness in Twitter can precede for a few hours and is highly unstable compared to the one in web. The application of these findings is also discussed on ad keywords planning in Search Engine Marketing.

The development of microblog services has a considerable effect on the patterns of web access and Internet resources discovery.Understanding the interrelation between information diffusion in ordine social media and user web interests can help the web ecosystem stakeholders in developing new services and designing efficient systems with optimized resources.This paper explores whether or not one can infer the trends of topics in the web by observing the Twitter microcosm.Using datasets collected from Twitter and two representative web services(Google and Alexa),this work conducts a comparative analysis between trending patterns of topics in Twitter and in the web by considering both the temporal and spatial perspectives,and finds that individual topics in Twitter and in the web share similar trending patterns both from the temporal and spatial aspects.Nevertheless,the trendiness in Twitter can precede for a few hours and is highly unstable compared to the one in web.The application of these findings is also discussed on ad keywords planning in Search Engine Marketing.