Turkish Language Association launches massive dictionary and AI model to boost Turkish usage

Osman Mert, head of the Turkish Language Association (TDK), outlined a suite of initiatives to modernise Turkish language resources. The flagship effort is a new dictionary projected to contain over 800,000 words, far surpassing one outlet 132,000-entry edition, and will draw from a balanced corpus of one billion words. Parallel to this, TDK has spent three years building a large-language model designed to let artificial intelligence process and generate text in Turkish; the theoretical phase is complete and practical work, including extensive data labeling, has begun.

The agency also plans to open its Türkiye Terminology Bank, now holding about 310,000 terms, to the public and eventually expand it into a Turkic World Terminology Bank. Three software applications are being prepared to automatically spot grammatical, spelling and punctuation errors, monitor public-sector websites for language mistakes, and flag words missing from the dictionary, with all expected to be operational within a year. Educational programmes for teachers and university students are being broadened nationwide to promote accurate Turkish usage in the digital age.