Introduction
Welcome to the user manual of EI-Authoring.
The authoring application is a workspace with an intuitive Web UI where users—data scientists, knowledge engineers, data analysts—, alone or in teams, can:
- Train, tune and deploy predictive language models that can either categorize/classify texts or extract information from texts.
- Manage thesauri with a powerful editor, with the possibility of creating predictive language models capable of extracting the citations of thesaurus concepts from texts.
- Create corpora of documents, for a discovery activity as an end in itself or to reuse corpora as libraries—for training or testing purposes—in other projects.
- Explore the knowledge graphs for the available languages, to make discoveries that are useful for other project activities.
- Create customized versions of a knowledge graph starting from the factory version.
EI-Flow is the complementary application that makes models usable via an API by putting them inside document analysis workflows.
This manual is divided into two sections:
- How-to guide: all the procedures to use the application.
- Reference: definitions, lookup lists, formats