Innovative Monitoring Systems and Prevention Policies of Online Hate Speech
Sector: Education • Location: Slovenia, Italy, Cyprus, Belgium
Source: EU Funding & Tenders Portal
Hate speech is spreading not only on social media, but also in the comments of online media news sites. Different ways of presenting news trigger different, sometimes inappropriate, responses from target audiences. The objectives of the project are to automatically detect hate speech in several languages, to identify triggers, to recommend effective counter narratives, and to propose policies to t
Project Information FAQ
Project Information
Want to explore the full details? View the full report
Participants
Sponsoring Agency | Obfuscated Data |
Company | Obfuscated Data |
Status
Original status | ended |
Taiyo status | Obfuscated Data |
Taiyo last update | 00-00-0000 |
Available timestamps | 00-00-0000 |
Available timestamp type | Obfuscated Data |
Contact
Contact name | Obfuscated Data |
Phone | 0000000000 |
ObfuscatedData@email.com | |
Address | Obfuscated Data, Obfuscated data, obfuscated data, Obfuscated data |
Description
Description | Hate speech is spreading not only on social media, but also in the comments of online media news sites. Different ways of presenting news trigger different, sometimes inappropriate, responses from target audiences. The objectives of the project are to automatically detect hate speech in several languages, to identify triggers, to recommend effective counter narratives, and to propose policies to the European communication regulators. The consortium consists of top research teams in the areas of Artificial Intelligence and machine learning (JSI), misinformation spreading (UNIVE), and hate speech counter narratives (UCY). Effective deployment and long-terms sustainability will be ensured by an experienced SME (Textgain). The policies proposed will be disseminated to the European Regulators Group (ERGA), where the Italian communication regulator (AGCOM) is heading a task force. The technical part of the project is data-driven. We closely collaborate with two prime national media and collect online content in Italian (Corriere della Sera) and Slovenian (RTV SLO public TV). We will also include English and Dutch social media contents from Twitter, Instagram, and Reddit. The data will be manually annotated for types of hate speech, and quality carefully controlled. We will develop hate speech models by applying advanced machine learning methods. The models will be deployed in online settings and used to analyze the large amounts of data already collected. This will allow to analyze mechanisms and trends in hate speech, and identify different types of triggers. We will conduct a pilot study on possible counter narratives and develop recommendations. The dissemination activities of the project will target European regulators (ERGA), online media companies and journalists, and the scientific community. We will organize training activities for civil servants and journalists. There will be a communication campaign to raise awareness among the general public and youth. |
Original sub-sector | Obfuscated |
Original Currency | USD |
Original budget | 000000000000000 |
Procurement method | Obfuscated Data |
Budget | 000000000000000 |
Location
Region | Obfuscated |
Country | Obfuscated |
State | Obfuscated Data |
County | Obfuscated |
Location | Obfuscated Data, Obfuscated data, obfuscated data, Obfuscated data |
Source
Source reliability | High |
Data quality score | 100% |
Source | Obfuscated Data |
URL | obfuscated_data,obfuscateddata.com |
More Details
Project Type | Obfuscated Data |
Article Published Date | Obfuscated Data |
