Turning Linguistic Assets into Knowledge for GenAI Applications

Companies struggle with getting GenAI into production. And well-structured, semantically enriched data can significantly improve the reliability and control of content creation and also prepare content for AI. 

This is good news for the localization industry, because our linguistic assets like terminologies play a vital role in that! On the one hand, terminology can directly provide guard rails for AI-based content generation and translation and make content more reliable by means of Terminology Augmented Generation (TAG). On the other hand, terminology can feed data into taxonomies and knowledge graphs, which in turn can enhance content delivery, search and retrieval, or even make recommendations on contents by means of Graph Retrieval Augmented Generation (GraphRAG). 

This webinar digs deep into the underlying use cases and technologies and tries to establish the difference but also the immense synergies between terminologies, taxonomies and knowledge graphs. We will show a lot of live uses cases:
 
•    What can you do with terminology for linguistic use cases in GenAI. We will show some live examples of Terminology Augmented Translation for terminology checking, translation but also text generation. And we will show why the TAG approach is more precise than the often-used RAG approach.
•    Terminology alone can also already help for question answering, at a similar level as RAG. There are limits to that, though. But even so, we will show that terminology is a great basis to build a taxonomy or knowledge graph and integrating the two makes sense.
•    With knowledge graphs, you can get much higher accuracy for question answering, for example with chatbots. But of course you can take knowledge graphs much further for use cases like auto-classification, recommender systems etc. 
•    The bottom line is: It depends strongly on the use case whether you want to use pure terminology or go deep into knowledge graphs. In either case it makes sense to combine the two to save time and reduce redundancies.
 

Audience

-    Localization managers
-    Terminologists 
-    AI specialists
 

Host organization: Kaleidoscope GmbH