We are proud to announce a groundbreaking achievement in the field of natural language processing (NLP) and sentiment analysis for the Bulgarian language. The pioneering SM-FEEL-BG project has successfully created the first Bulgarian datasets and classifiers for detecting emotions and sentiments in social media texts.
This research was led by the team at GATE Institute, including Irina Temnikova, Ruslana Margova, Silvia Gargova, and Nevena Grigorova, working closely with Identrics’ language technology team throughout.
What is the SM-FEEL-BG project?
The SM-FEEL-BG project focuses on developing large-scale datasets and advanced classifiers for the Bulgarian language. These tools detect and analyse emotions and sentiments in social media texts, filling a real gap in Bulgarian-language NLP.
This project marks a leap forward in natural language processing for Bulgarian, addressing a previously unmet need for localised tools and resources.
By creating the first datasets and classifiers of this kind, SM-FEEL-BG lays the groundwork for future work in countering disinformation and propaganda in Bulgaria, social media analysis, customer service, and public opinion monitoring.
Why is the SM-FEEL-BG project so significant?
For the Bulgarian tech community
SM-FEEL-BG gives researchers and developers access to large-scale datasets and sentiment classifiers built for the Bulgarian language, for the first time, opening up new ground for research and development in Bulgarian NLP.
The open-access approach to most of the datasets encourages a collaborative research environment, and shared knowledge among professionals in the field. It places Bulgaria on the global AI and NLP map, and gives the local tech community the tools and resources needed to drive future work, including the potential for Bulgarian language technology to compete internationally.
For businesses operating in Bulgaria
Businesses can use these tools to gain deeper insight into customer emotions and sentiments, improving customer engagement, service delivery, and how well their offerings meet audience needs.
The project also plays a real part in combating disinformation and harmful content on media platforms. By accurately detecting emotions and sentiments, these tools help identify and slow the spread of disinformation, hate speech, and other harmful content, contributing to a safer, more trustworthy online environment.
The significance for Identrics
This project reflects the expertise driving Identrics’ language technology work, and the depth of what our team builds when working alongside strong research partners like GATE Institute.
It sits well within our broader mission: using AI and machine learning to combat propaganda, analyse unstructured data, and automate work that would otherwise take analysts far longer to do by hand.
As we continue building AI-powered solutions for clients, the insight and technology from this project feed directly into our wider work, from Kaspian’s enrichment services to the models behind Sitreps and Annex.
Congratulations to the team
We congratulate the GATE Institute team for their work, and thank everyone at Identrics who contributed to the project. This kind of collaboration is what moves Bulgarian-language NLP forward.
At Identrics, we’re committed to putting advancements like this to work, building AI solutions that fit real business needs. Reach out if you’d like to learn more about how our AI solutions can help yours.
About the author
Identrics Editorial Team
Editorial team, Identrics
Reporting by the Identrics editorial team - covering company news, partnerships, conferences, and joint research. Reviewed by senior staff before publication.
Reviewed by
Chief Executive Officer, Identrics