Rumor Identification in Social Networks


Rumor is a type of unverified and instrumentally relevant information that is circulating in ambiguous contexts and that functions to help people make sense of their surroundings, manage risk or deal with perceived threats. It is usually spread by word of mouth, but may be transmitted by electronic means as well.

In social science, rumor has been studied for decades and has become an important research topic in the field of epidemics, sociology and psychology. It is widely spread and can ruin reputations when it is false or isn’t disproved quickly enough, and it is difficult to identify the source of a rumor.

Traditionally, rumors have been classified into four types: wish rumors, fear or bogey rumors, wedge-driving or aggressive rumors and anticipatory rumors. However, recent studies have shown that rumors also can include positive (pipe dream) and negative (bogie and wedge-driving) elements.

The most common rumor in everyday life is the news. The news is usually related to current events, but it can also be about past events. Therefore, the news is very interesting when it appears for the first time, but is not so interesting when the same event happens again.

It is not always easy to distinguish between news and rumor, and many people believe that a rumor is more trustworthy than a news story. In addition, a rumor is often more likely to be passed along and spread than a news story.

This study investigates the process of rumor spreading in social networks and proposes a model to identify rumor credibility. The model takes into account rumor credibility, the correlation between rumors and people’s lives and crowd classification based on personality.

A set of coding systems was applied to archived discussions concerning rumors on the Internet and other computer networks such as BITnet. The results showed that the coding system was effective in identifying the four stages of rumor discussion: prudent, apprehensive, authenticating, and interrogatory.

There are three important characteristics of rumors that should be considered in a rumor identification task: the correlation degree between rumors and people’s lives, the importance of the subject matter, and the ambiguity in the rumor. The correlation degree is one of the most significant factors in determining whether a rumor will be spread.

The importance of the subject matter is another important factor that influences whether a rumor will be spread. If the subject matter is not important, a rumor will not be spread. It is also important to consider the ambiguity of the rumor when determining whether it will be spread.

Moreover, a rumor that reveals an ambiguity will be more likely to be spread than a rumor that does not reveal any ambiguity. The rumor that reveals an ambiguity has the ability to attract attention and evoke emotion, as well as incite involvement.

We developed a new mathematical measure of rumor spread called Spread Power of Rumor (SPR). This is the first empirically-based measurement in the field of rumor analysis.