Monok at International Language Generation Conference

Monok founder Adam Sam participated as an industry panelist at the 2021 International Natural Language Generation Conference (INLG), hosted by the University of Aberdeen in Scotland.

The conference, held September 20–24, brought together international researchers and industry specialists to discuss advances in natural language generation (NLG) and its applications in commercial environments.

From research to real-world applications

The panel, titled "What Users Want from Real-World NLG!", explored the challenges of developing language generation technologies for practical, real-world use.

The discussion featured representatives from four companies:

• Adam Sam, Monok — Automated journalism and content production
• Ross Turner, Arria — Data-to-text generation
• Robert Weißgraeber, AX Semantics — Automated content generation
• Michelle Zhou, Juji — Conversational AI

The panel brought together perspectives from automated reporting, digital publishing, and conversational systems, examining how commercial requirements influence the development of language generation technologies.

Human expertise in automated journalism

Monok develops technology for automated news production, combining natural language processing, information extraction, and text generation to transform information from multiple sources into publishable articles.

During the panel, Sam explained how Monok leverages the work of its human writers as training data for its text generation systems, using editorial expertise to improve the quality of automated content.

Rather than treating human writers and automated generation as competing approaches, Monok combines them in a process where human-produced and edited content contributes to the ongoing development of its language models.

This approach connects the practical knowledge of professional writers with automated publishing technology, helping address challenges such as factual accuracy, linguistic quality, and consistency across different content categories.

The discussion highlighted the importance of human expertise in developing language generation systems that meet the requirements of professional publishing.