Q1 2023

Follow the project and be the first one to try out this unique API

The problem

Create a base of questions and answers to generate a dataset for training chatbots can be a laborious and often a longstanding process. Many companies find it difficult to allocate human resources to realize this task, which makes this process more complex, slower and much more costly than expected.

Unlike the known automation scripts, Got It AI uses exclusive proprietary technology that does not yet exist in the market, enabling automation of the FAQ creation process, and raising the quality standard of questions and answers to the human level.

The solution

To create assertive questions in an automated way, we use the concept invented by Aristotle in his main work, known as "Nicomachean Ethics", where he contemplates about the seven circumstances (Septem Circumstantiae) which we refer to: Who? What? Where? When? Why? How? By what means?

We expanded our vision with the journalistic concept known as LIDE (Who? What? When? Where? How? Why?), as well as the most current investigative concept known as 5W2H (Who? What? Where? When? Why? How? How much?).

See our ENGINE IN ACTION

Select one of the texts below to be analized

Analyze a text to see the result...

Q1 2023

Follow the project and be the first one to try out this unique API

How it Works

Based heavily on Aristotle's innovative concept, on the rereading LIDE and on the 5W2H method widely recognized in the corporative environment, we synthesized our engine, and our own concept.

Got It AI's proprietary artificial intelligence engine powers an all-new concept called 12W6H, which includes the questions: Who? What? When? How? Where? By whom? By what? For whom? For what? With whom? With what? According to Who? For what purpose? In What condition? How much? Why?

This much broader and specific set allows us to identify the information in a more precise and complete way, making it possible to create questions and answers for any query in relation to a content.

To do it the right way, our solution goes beyond the use of conventional concepts in NLU that identify entities and group them by categories, as well as goes beyond the concept of using vector of words to try to understand their meaning by statistical means.

Our solution mimics exactly the way human beings process written information: We use pure semantics, through the mathematization of grammatical, syntactic and morphological linguistic rules, known by all of us, but until now, not recognized by current computing.

The results are questions and answers that look like they were created by humans with great cognitive ability, SAID NO ONE EVER, they came from Got It AI.

Now, what really matters: CODE!

Above you could see how our engine works. Now see how easy it is to use our API to generate your Chatbot's training dataset, with just one simple RESTful request.

IMPORTANT: Our API also allows you to generate the ".TSV" file with the QnA pairs, ready for import!
REQUEST body  (JSON)
{
    "text": "Astronomers have spotted three near-Earth asteroids that were lurking undetected within the glare of the sun. One of the asteroids is the largest potentially hazardous object posing a risk to Earth to be discovered in the last eight years. Their findings were published Monday in The Astronomical Journal."
}
RESPONSE body  (JSON)
{
  "success": true,
  "result": [
    {
      "q": "WHO have spotted three near-Earth asteroids that were lurking undetected within the glare of the sun?",
      "a": "Astronomers"
    },
    {
      "q": "WHAT did astronomers have spot lurking undetected within the glare of the sun?",
      "a": "three near-Earth asteroids"
    },
    {
      "q": "WHO is the largest potentially hazardous object posing a risk to Earth?",
      "a": "One of the asteroids"
    },
    {
      "q": "WHAT One of the asteroids is?",
      "a": "the largest potentially hazardous object posing a risk to Earth"
    },
    {
      "q": "WHAT is One of the asteroids?",
      "a": "the largest potentially hazardous object posing a risk to Earth"
    },
    {
      "q": "WHEN were Their findings published in The Astronomical Journal?",
      "a": "Monday"
    },
    {
      "q": "WHO were published Monday in The Astronomical Journal?",
      "a": "Their findings"
    },
    {
      "q": "WHERE were Their findings published Monday?",
      "a": "in The Astronomical Journal"
    }]
}

Q1 2023

Follow the project and be the first one to try out this unique API