Competentum AI for Automated STEM Content Generation


Accepting the challenge to reduce the high costs of creating quality content, Competentum takes an automated approach to STEM problem development by applying its expertise in artificial intelligence and machine learning. We are building an innovative platform, EdWise, that will allow automated generation of personalized adaptive content for various STEM disciplines.

Machine Learning for Natural Language Processing

Machine Learning for Natural Language Processing (NLP)

EdWise uses deep neural networks and other machine learning methods for natural language processing. NLP models like Word Embeddings, Topic Modeling, and use of Thematic Ontologies enable us to train an AI-based machine to process data in the natural language and generate new high-quality content more quickly, in greater quantity, and less expensively than a fully human expert-based approach.

Competentum EdWise Platform

EdWise Modules for Classifying Datasets & New Content Generation

The two main components of the EdWise platform are the classification and generation modules. The classification module links an original set of tasks to a particular taxonomy of a certain subject area. The generation module creates ontological spaces and generative grammars from the original datasets. When a request for generation is received, EdWise classifies it, determines the necessary generative grammar and ontology, and creates a new task.

Competentum Personalized and Adaptive Content

Personalized & Adaptive Content at Scale

Personalized and adaptive content increases the effectiveness of teaching and improves the learning process. EdWise creates ontological spaces depending on a student profile, learning objectives, and areas of interests, which allows the generation of personalized tasks.

A large number of questions linked to certain taxonomies and generation of “similar” tasks enables EdWise to adapt the content for individual learning paths. Since this process is automated, we can deliver personalized, adaptive content at scale.