Web event is a story or a scandal occurred in the society or on the web reflected by a series of associated web pages with time. It is hard to measure web event veracity as the result that volume and various web events happen all the time. However, veracity of web event has to be detected in order to get trustworthy and valuable information. In its evolution process, social event is deeply affected by corresponding web information, whose one performance is uncertainty. In other words, monitoring veracity of web event via uncertainty can help user understand social events. Generally, event with high uncertainty are more likely to turn into popular or emergent event (Haddow et al., 2010). Here uncertainty of web event is determines by its features distribution on different kind of confident websites instead of entropy used in informatics. For instance, if an event has high uncertain distributed in low confident website, then the event is more likely to be faked; namely, its veracity is low; and vice versa. So, this research employs multi-factor based uncertainty to measure the veracity of web event. However, it is difficult to manually analyze the veracity of the volume web events in their evolution process on the web, because it is a killing of time and energy. So, to understand and quickly respond to the volume web events, it is necessary to measure the web event veracity automatically.