Update README.md

README.md

@@ -7,4 +7,64 @@ sdk: static
 pinned: false
 ---
 
-Edit this `README.md` markdown file to author your organization card.
+# FactDeMice: Evidence-based Fact-Checking with Fact-consistent translation, Fake Review Detection, and Automatic Misinformative Claim Extraction
+
+
+In today’s digital world, public opinion is shaped largely through online communication—whether it’s news articles, comment sections, or product reviews. Unfortunately, these same channels are also exploited to spread misinformation and fake reviews, influencing markets, public trust, and even democratic processes. The **FactDeMice Project** (Fact-Checking and Detection of Misinformation in Comments and Reviews) aims to address this urgent societal challenge by developing a robust, AI-driven toolkit to detect, analyze, and counter false or misleading information across languages and media platforms.
+
+---
+
+## Project Goal
+
+The overall goal of FactDeMice is to create a **FactDeMice Toolkit**—a comprehensive software suite that empowers journalists, analysts, and organizations to identify misinformation and ensure the accuracy of public information. The toolkit combines cutting-edge artificial intelligence, machine learning, and natural language processing methods with fact-checked databases and advanced translation systems to support reliable, multilingual fact-checking.
+
+---
+
+## Key Objectives
+
+1. **Detection of Fake Reviews:**  
+   Identify deceptive reviews using stylistic, statistical, and behavioral analysis of online content and user profiles.
+
+2. **Detection of Disinformation Campaigns:**  
+   Recognize orchestrated online efforts that aim to manipulate public opinion through coordinated false narratives.
+
+3. **Automatic Claim Extraction:**  
+   Extract and verbalize factual claims from news comment sections to enable systematic fact-checking.
+
+4. **Evidence-Based Verification:**  
+   Retrieve supporting and refuting evidence for claims and generate faithful, contrastive summaries highlighting both perspectives.
+
+5. **Factually Consistent Translation:**  
+   Develop translation systems that maintain factual consistency across languages (English, Czech, and Mandarin), enhanced through **Reinforcement Learning with AI Feedback (RLAIF)**.
+
+
+## Expected Results
+
+By the end of the project, the consortium will deliver:
+
+- **FactDeMice Toolbox (D3.1):** A comprehensive misinformation-detection software toolkit.  
+- **Consistent Translator (D3.2):** A machine translation system ensuring cross-lingual factual consistency.  
+- **Public Datasets (D2.1–D2.4):** Annotated datasets of fake reviews, misinformation incidents, target claims, and verified evidence sources.
+
+---
+
+## Partners and Collaboration
+
+The project unites leading European and Asian research and industrial partners:
+
+- **LINGEA (Czech Republic)** – Project coordination, integration, and commercialization.  
+- **Brno University of Technology (BUT)** – Research on evidence retrieval and contrastive explanations.  
+- **National Taipei University of Technology (TAIPEITECH)** – Research on fake-review detection and analysis.  
+- **Institute for Information Industry (III, Taiwan)** – Data collection and AI development support.
+
+---
+
+## Business and Societal Impact
+
+**Business impact:**  
+FactDeMice creates new opportunities for media companies, e-commerce platforms, and organizations seeking to protect their reputation and ensure information integrity. Its commercialization potential includes integration into Lingea’s professional software services and as a public fact-checking web platform.
+
+**Societal impact:**  
+By combating fake news and disinformation, FactDeMice promotes transparency, critical thinking, and public trust in information ecosystems. It supports democratic resilience, informed decision-making, and fair competition in online markets.
+
+---