KIMO uses a variety of technology at the back-end, the most important elements you can find below with a brief explanation. Also check-out our Github profile where we share the open-source code available.

  • Natural Language Processing (NLP): interacting with the chatbot in KIMO is based on a technology called NLP. NLP is technically a system that allows a computer to predict the right response to an input, were that input consists of human language. It enables a different way of communicating with a machine.
  • Deep Learning (DL): user preferences are stored in patterns through a technique called Deep Learning. This system uses ‘neural nets’ to essentially store patterns, relevant features and make predictions. Note that Deep Learning is a subset of the AI-algorithms that is predicted to generate 40% of the expected value in AI.
  • Unsupervised Learning (UL): KIMO uses UL algorithms to find similarities between content or users. These similarity scores are then used to make predictions for the right content per user. UL is used to cluster and predict based on unstructured (unlabeled) data.
  • Sequence Mining (SM): also called sequential pattern mining, algorithms in this space are focused on creating sequential patterns in data. In KIMOs case, this is used to track learning paths for people (e.g. video 1 -> video 3 for user X).