Assets

Linguistic Assets

Congree UCC is used to develop sustainable assets that help define your corporate identity, manage knowledge, and train AI systems. Congree UCC's tools make all of these assets accessible throughout the company, and they can be shared with other processes by way of programming interfaces or standard export formats.

Treasures Lie Dormant in Texts
Automated Treasure Hunt Employing Linguistics and Artificial Intelligence

Manually compiling corporate terminology can be a very tedious and time-consuming task. But it doesn't have to be that way. Linguistics and AI can easily analyze large amounts of data in Congree UCC and do the groundwork for experts in the company. This dramatically speeds up terminology work and unearths hidden linguistic treasures.

Of course, linguistics and AI can also immediately recognize term candidates or knowledge relations while content is being created and propose them to the author.

Terminology and Corporate Identity

Terminology is crucial for corporate identity. Specific terms demonstrate expertise and position your company as an industry expert. The unique use of language can convey the company's values and set it apart from competitors. Well thought-out terminology thus strengthens corporate identity and cements the company's image.

Terminology as a Multilingual Knowledge Graph

Terminology is the embodiment of corporate knowledge in a compact and structured form. The terms, definitions, and relations employed constitute high-quality, verified knowledge. The term-based approach creates a multilingual knowledge graph illustrating internal relationships, which in Congree UCC is available to all employees via tools such as Authoring Assistance or Congree TermTiger.

Using Linguistic Assets to Train Language Models

Adapting a generic language model requires company-specific data. The assets stored in Congree UCC provide an excellent basis:

  • The terms in the Terminology largely define your corporate language.
  • The rules and examples in the Style Guide Manager determine the wording required in specific contexts.
  • The Authoring Memory includes every approved sentence and offers an extensive corpus of positive examples for your corporate language.
  • Anonymized Reporting can collect negative examples in Congree UCC along with text corrections. These corrections can be taught to the AI. 

Linguistic assets allow you to create sustainable value and make you less dependent on AI providers. You are always prepared to adapt generic language models with good data.