Comment on page
Why Mind Expression
- Mind Expression is a Conversational AI as a Service (AIaaS) for the development of Conversational AIs (also known as "chatbots").
- With a context-aware development interface, a developer can easily create complex conversation flows and test them immediately in the Sandbox.
- With a single, simple public API access, integration with any front end is easily interfaced.
- Developers can easily add specific knowledge from their own organizations and use it immediately.
- Unlike every other chatbot platform, our Conversational AI Builder does not rely on a machine learning (ML) framework to properly match a chatbot user's "intent".
- Mind uses a new Symbolic Paradigm, which overcomes the problems of the older, first-generation symbolic models, solving the problems of brittleness and adaptive learning.
- Mind doesn't rely on Big Data; we do not train our model, we educate it.
- Neural network (NN) and Machine learning (ML) models are not made to reason: they only spot patterns, and they don't know why something made "sense" to them or not, whereas Mind understands what's going on and why.
- Since we are not dependent on training phrases, the ease of setup is unparalleled, and the development time is only a fraction of the ML models used by everyone else.
- Other systems use training phrases (one after another after another) to train their "intents", whereas Mind just requires one logical statement to teach the meaning of what the developer wants to handle.
- As the components and actions that make up the flows are available in a context-aware development environment, the developer can easily add what is logical at every step of the building process.
- Interfaces to enterprise back-ends & fulfillment are easily configured without the need for any programming skills to develop them.
- Transparency: it is well known that ML/statistical/NN implementations are "black box" technologies.
- In contrast, the primary model we utilize, the Canonical, is modeled to be understood by both a human being as well as a machine, so transparency is available at every level.
- Our model is a Cognitive model: the phrases are understood logically and are logically related to other parts of the ontology by their meanings.
- In other words, the other systems only approximate meaning, whereas we deal with actual meaning, that makes sense to everyone.
- No other platform gives the capability of Context Hopping, which is available by default in any conversation flow created by the developer: no extra setup is required.
- What is available in other platforms is a set of very rudimentary connections that are arbitrarily connected, and whose context is lost unless specifically programmed into the chatbot model that is being developed.
- Hopping means jumping from point to point in a Subject and between Subjects, where no context is lost in between the "hops".
- Where there is more context needed to properly enter a new state, we automatically go to where the required information is collected.
- Another related feature is "arbitrary entry points", which are like hopping into a flow from nowhere; these are also provided without additional setup.
- Since everything is logically understood, the transfer of knowledge is built-in.
- Between domains, an ontology that is understood by a certain context can be translated into other contexts for different domains (analogical reasoning, based on the cognitive model of abductive, deductive, and inductive reasoning).
- Between languages, as we work with meanings, what works in one language will work the same way in another language.
- Along with the ease of setup, what is built is more accurate, too, than the ML/NN models available.
- You can easily insert new states where they logically fit (guided by our context-aware UI), and you don't have to worry if you've done something wrong in the setup.
- All changes cascade so the new flow is assimilated without any hand-rewiring.
- Context hopping is automatic, and it does not need to be configured, because everything has a specific purpose, and the context is kept throughout the conversation.
Last modified 1yr ago