On the 5th of November 2020, researchers of the Artificial Intelligence National Laboratory held an online webinar in front of scientists and industry partners to discuss challenges faced in language technology and automation of customer and other communication driven services. Such challenges may be solved by the extensive use of AI yet teaching machines to understand idioms is no easy task.
Gergely Szertics – Language transformation
Gergely Szertics, all-time consultant of Hungary’s Digital Success Programme presented about the complexity of language transformation in automation services operating in Hungarian. The presentation exemplified the main modules and actual fields of use of the national AI Strategy’s referenced project.
Ádám Feldmann – Creating large scale AI models in cloud environment
Ádám Feldmann, senior lecturer of the University of Pécs, presented a pipeline that may help to implement large scale language technology models within Hungarian environments. Due to the fact that currently available language models’ are mostly function as frameworks, feature-driven development requires personalizes infrastructures.
Richard Farkas – Applicable semantical and hybrid intelligence research within the Artificial Intelligence National Laboratory
Richárd Farkas, senior researcher of the University of Szeged introduced the audience to the current language technology applications at the National Laboratory, while also explained the required researches to achieve actual, widespread industry use of such applications. Challenges include the translation of language semantics to computer code and the specializations of deep learning algorithms to different fields of automation.