What powers Vecgraph voice apps?
Vecgraph voice apps are incredibly intelligent. We developed our AI specifically to power our explainable content recommendations. This allows our voice apps to identify more relevant content, and also explain to the user why they might be interested. Not only do our voice apps use Natural language understanding (NLU) to understand what your users mean, but they also use state of the art machine learning techniques to automatically understand and recommend your content.
AI to understand your content
Your voice app is powered by a content graph which is a representation of the relationships between your content. Your voice app uses this graph to recommend relevant content to your audience. There are several techniques used in creating this content graph.
Named Entity Recognition (NER)
Vecgraph uses NER to extract entities from your content. These entities include names, organizations, and places. This allows our voice apps to explain to the user why they might be interested in another story. The extracted entities allow our AI to recognize the relations between content by recognizing which content mentions the same or similar entities.
Vecgraph disambiguates entities (such as people’s names or organizations) and links the entity to its data in a knowledge base. This data is then used to identify the relations between the extracted entities. Some examples of these relations are, athletes identified as being on the same team, or one company identified as being a subsidiary of another. These relations are used by our AI to identify relations between content.
Natural Language understanding
NLU allows our voice apps to actually understand what the user is saying. It extracts key phrases and makes sense of the conversational context. We use NLU to make the user’s interaction with the voice app as natural as possible. We strive to create voice apps that deliver exactly what the user is looking for. By allowing the user to interact using natural language we improve the ease of use and avoid the dreaded “Sorry, I didn’t understand that, can you try again?”.