There is still a relative lack of consensus in the academic literature on key terms and concepts used in disinformation research. In the literature review we discuss some of the concepts, theories and thematic debates in more length. In the interest of brevity and to aid our analysis, our project uses working definitions of key concepts. These are as follows:
Artificial Intelligence (AI)
Broad term, generally used to describe any machine learning or neural network based algorithm than can perform task on new data based on previous learned patterns.
Content Analysis
The methodical examination of media and communication content to understand and interpret its meaning.
Coordinated inauthentic behavior (COO)
Fake accounts, groups or pages aimed at misleading or deceiving users in a systematic fashion
Data Collection
The process of gathering information from various sources for analysis.
Evaluation metrics
Metrics to evaluate the performance of an algorithm by comparing its output to a ground truth.
Disinformation
Intentional spread of misleading or false information, often for political gain or to manipulate public opinion, with harmful effects.
Information Campaigns
Coordinated efforts to disseminate specific information or propaganda to influence public opinion.
Information Disorder
The combination of disinformation, misinformation and mal-information which undermines the veracity and integrity of information in the public sphere
Low-resource languages
Low-resource are those languages that have less data available for training conversational AI systems. Arabic is a low-resource language. English, Chinese and French for example are high-resource languages and provide more data for AI experiments.
Inter-coder Reliability Test
A method to measure consistency among multiple coders in how they categorize or code data.
Malinformation
The use of information which may not necessarily false, but used out of context or without permission, to cause harm, e.g. the publication of private information (doxing or revenge porn) to exert pressure, shame or cause harm
Media Narratives
The overarching stories and messages conveyed through media content.
MIL
Media and Information Literacy. The ability to access, analyze, create and reflect on media. It is a prerequisite for citizens to exercise and claim their rights to freedom of expression and access to information
Misinformation
Inaccurate information which may be harmful or misleading but is spread without the intention to cause harm
Propaganda Analysis
The examination of propaganda to understand its techniques, messages, and impact.
Social Media Platforms
Online platforms where users create, share, or exchange information and ideas in virtual communities.
Sputnik Arabic
Arabic version of Sputnik, a Russian state-owned news agency and radio service.
Data Analysis
The process of examining, cleaning, transforming, and modelling data to discover useful information
Natural Language Processing (NLP)
Natural Language Processing (NLP) is an advanced technology that employs artificial intelligence algorithms to analyze large volumes of text data, enabling the identification of patterns and themes.
Content Dissemination
The process of distributing information or media content to a target audience.
Data Analysis
The process of examining, cleaning, transforming, and modeling data to discover useful information.
Information Consumption
The process by which individuals access, interpret, and use information.
Information Warfare
The use of information and communication technologies by a state or organization for hostile purposes.
Fact-checking
The process of checking that all facts in a piece of writing, a news article, video etc. are correct.
FIMI
Foreign Information Manipulation and Interference describes a mostly non-illegal pattern of behaviour that threatens or has the potential to negatively impact values, procedures and political processes. Such activity is manipulative in character, conducted in an intentional and coordinated manner, by state or non-state actors, including their proxies inside and outside of their own territory.
Machine Learning Models
Computational models that use algorithms to analyze data and learn from it, often used in AI.
LLM
Large Language Models are computational models capable of understanding and generating natural language. They learn from vast amounts of data and can perform a wide range of language tasks.
Manual Coding
The process of categorizing data by human analysts based on predefined criteria.
Misleading Information
False or inaccurate information that can deceive the intended audience.
Propaganda
Strategic dissemination of information to support specific political agendas or narratives. It may contain truths but is attempting to persuade the public to support a particular, often malevolent, point of view, and is therefore usually associated with negative outcomes.
RT Arabic
Arabic-language channel of Russia Today, a Russian state-funded international television network.
Social Networks
Platforms and communities where people interact and share information online.
State-Sponsored Media
Media outlets that are funded and controlled by a government, used to disseminate its perspective.
Telegram
A messaging app popular in the Middle East and Russian contexts, used in this project for sourcing data.
