The Web has shown that making information (of any kind) widely available to everybody has positively impacted the way people live: how they work (e.g., affecting the industrial world at any level) and how they conduct daily, cultural, social, leisure and other activities. However, a large part of the information comes from sources that we are not able to verify. This prevents us from utilizing it, as we are simply very uncertain about it. The most recent example of such information is contained in social streams, such as Twitter, Facebook, YouTube, etc. These networks provide a huge amount of data, which, thanks to its dynamic online nature, covers very recent facts and trends, often presenting some information earlier than traditional media. Unfortunately, such content is contaminated by false statements, accidentally or deliberately made by users. Again, there is no means to verify such information, we can only try to follow the structure of facts, e.g., related stories, by performing time consuming query search, with the hope of finding positive or negative evidence of the claim’s veracity.

A more critical problem relates to the claims made by prominent politicians or other important society players. Often, citizens cannot make the appropriate decision because they are presented with biased and even false information. This important problem is tackled by manual fact-checking organizations like factcheck.org and PolitiFact.com, who aim at provide a grade on important statements thus contributing to the formation of a political opinion.Fact-checking is an urgent need of our society and manual fact-checking proves the humanity’s desperate need for it: billions of facts on the web cannot be checked by hand. Unfortunately, there is no technology enabling automatic fact checking: search engines, such as Google, Bing, Yahoo! cannot perform the kind of deep semantic analysis and fact correlation that is required for this high level semantic task.

However, the experience of the famous IBM Watson system has demonstrated that high accurate semantic processing can be carried out for automatically deriving answer with accuracy (and of course speed) higher than the best human experts.This suggests that we have the means to develop a fact-checking technology.In particular, the development of such technology will require a joint interdisciplinary effort among: (i) Information Retrieval, for developing efficient access to many different sources of information, (ii) Computational Linguistics for analyzing the semantic content of facts, (iii) Machine Learning for learning to carry out inference on the acquired semantic information, and (iv) social and political science, which will provide the structure and foundations for combining the above different techniques in effective systems.

This technology has the potential of changing the way our society accesses online knowledge. By providing robust models for automatically identifying the veracity of facts in the realtime, it can improve the quality of information presented to the society and prevent manipulative online behavior. It can help citizens separate truth from fiction, e.g., in political discourse. This can have a measurable effect on politicians, journalists so that they become less likely to disseminate false and misleading information in the first place.

Finally, in addition to an active community in Europe, e.g.,

https://fullfact.org/europe/http://factcheckeu.org

USA is giving a lot of importance to the problem of fact veracity, and they have already started the discussion on what is needed to develop such technology as a recent conference has shown

http://www.poynter.org/2016/whats-does-the-future-of-automated-fact-chec...