X International conference�“Information Technology and Implementation” (IT&I-2023)�Kyiv, Ukraine

1

Dedicated to the tenth anniversary of the Faculty of Information Technology

Information System for

the Fact-checker Support

Information System for the Fact-checker Support

• In a world where inaccurate news can spread quickly and have harmful impact, the process of fact-checking is becoming vital and increasingly topical. It is a key tool in the fight against so-called fake news and disinformation. Fact-checking is not only a way to identify false statements but also a means of restoring the trust to the media.
• In this context, the development of an automated fact-checking tool can be not only a significant contribution to the field but also a real asset for journalists, fact-checkers, and anyone working in the informational and media sectors. It is believed that these kind of tools can greatly simplify the fact-checking process, reduce the time required for it, and increase the efficiency and accuracy of the process.

What is a fake?

• Fake news is fake information disseminated under the guise of news.
• The purpose of fakes is:
• influence on public opinion
• manipulation of society
• achieving certain political, economic or other goals
• spreading panic
• public outrage
• distortion of facts
• influencing the electoral process
• stimulating conflicts.

Fact-checking

• Fact-checking is a methodical process of identifying and verifying the truthfulness and accuracy of information.

​

• This procedure is important in the media industry, where journalists, editors and other professionals are required to effectively verify information before it is disclosed.
• Fact-checking includes checking aspects such as factual data, quotes, sources of information and other content.

Methods and Techniques of Fact-Checking

• Initial verification of facts
• Checking the reliability of information sources
• Checking the context
• Comparison of data
• Involvement of experts
• Using technology for fact-checking

Fact-checking services

• VoxUkraine is a fact-checking project of the independent analytical platform Vox Ukraine. The team exposes lies, manipulations and Russian propaganda in Ukraine and abroad.
• Stopfake.org - the main activity is aimed at combating anti-Ukrainian propaganda, statements and facts aimed at discrediting Ukraine.
• Gvara Media is an independent online publication about social change.
• FactCheck.org - a non-profit website that aims to reduce deception and misunderstanding in US politics
• Fact Check Explorer - a tool that shows the latest fact checks and allows you to verify information.

Information system architecture

Algorithm of the visualization module

�No Yes

​

Fuzzy text matching

• Cleaning of stop words: prepositions, conjunctions, pronouns.
• Cleaning of invalid data: links, etc.
• Create text embeddings using pretrained neural network
• Universal Sentence Encoder Multilingual model from Google
• Tensorflow library
• Each text represent as vector with flot 512 elements

​

​

​

​

​

​

• Apply cosine similarity as matching measure

​

​

​

​

​

Algorithm of the database search module

Algorithm of the Telegram search module

Telegram comments parsing algorithm

Algorithm of the analysis and verification module

True news:

• The news comes from a well-known and reliable source with a good reputation and credibility.
• There is an author of the article.
• The news has a professional appearance, grammatically correct style and a minimum number of errors.
• The news does not contain obvious signs of fake news or manipulation.
• The information is presented in compliance with journalistic standards.

​

News with suspicion of fake:

• The source of the news may be questionable or unknown, but is not completely unreliable.
• There is no author of the article.
• The news may contain some errors, typos, or unclear style, but overall appears acceptable.
• The news may have a sensational headline, use certain emotional

Rules for analysis and verification

• If the news source is unreliable, the value is 1, otherwise 0.
• If there are errors in the text of the news - value 1, otherwise - 0.
• If there are manipulative words in the text - value 1, otherwise - 0.
• If there is no author of the news, the value is 1, otherwise 0.

​

• We use weights for each rule. In our case, each rule has a weight of 0.25.The value is calculated by the formula: fake_news_score = (source weight * source rule + text weight * text rule + manipulation weight * manipulation rule + author weight * author rule).

Database search result

Search result in Telegram

Search and verification results in the network

Conclusion and future works�

• In this article, we have presented a comprehensive system for fact-checking and analyzing news content, which utilizes the power of machine learning, natural language processing, and data integration.
• The effectiveness of our system is based on its ability to automate and expedite tasks that typically demand hours of manual fact-checking.
• One of the key strengths of our system is its integration with machine learning techniques, including state-of-the-art NLP transformers. We are currently exploring the development of a module for automatic text classification using these advanced technologies.

Thank you for your attention!!!