Summary

In this code pattern, find out how to utilize IBM ® Watson ™ Knowledge Studio to train a custom-made machine learning model to drive a decision-making procedure of identifying dangerous scenarios.

Description

Want to develop an application or service that can minimize the response time of very first responders? This code pattern describes how to develop a danger response web app that can help make choices quickly. For instance, when someone requires help, they call 911 and begin a dialogue with a dispatcher. The Watson Threat Response application screens the conversation for choose threats, including fires and floods. If the threat is identified as one of these, the suitable authorities look out immediately, which can lower the action time. Designers can build on this application to not just use the conversation, but also consist of social media posts to further confirm the situation. The application can be broadened to other domains such as schools and workplaces.

Flow

Build a Watson Danger response tool flow

  1. Audio input is captured by the Watson Speech to Text service.
  2. After being transcribed, the text is sent to the Watson Natural Language Comprehending service.
  3. Within the Watson Natural Language Comprehending service, a custom-made Watson Knowledge Studio maker finding out design processes the text for danger.
  4. The threat analysis from the machine finding out model is then sent out to the Watson Natural Language Comprehending service.
  5. The last output of the Watson Natural Language Understanding service is presented, and the decision is made whether to notify the authorities.

Instructions

Get comprehensive actions in the readme file. Those steps show how to:

  1. Develop and configure the Watson services.
  2. Release the application server.
  3. Deploy the application locally.