Speech Technology Magazine

 

Q&A: Dahl and Normandin Explore Conversational Technology Platforms

At the 2019 SpeechTEK conference Yves Normandin of Nu Echo, Inc. and Deborah Dahl of Conversational Technologies, will present "A Comprehensive Guide to Technologies for Conversational Systems." Conference chair Jim Larson talked to Normandin and Dahl to get a sneak peek of the session, and learn about conversational system technologies.
Posted Apr 5, 2019
Page1 of 1
Bookmark and Share

At the 2019 SpeechTEK conference Yves Normandin of Nu Echo, Inc. and Deborah Dahl of Conversational Technologies, will present “A Comprehensive Guide to Technologies for Conversational Systems.” Conference chair Jim Larson talked to Normandin and Dahl to get a sneak peek of the session and learn about conversational system technologies.

Q: Why is choosing a platform critical for developing natural language conversational systems?

A: Choosing the right platform is critical because if the wrong platform is chosen and the application doesn’t scale or the developers aren’t able to implement important features, this can lead to tremendous amounts of rework. Some examples of rework include redoing the dialog flow, redesigning the entities structure, re-annotation of data, and recoding the back end integration. If it isn’t possible to do the rework for reasons of schedule or budget, this can lead to an application with compromises in functionality or even to abandoning the application. This can be avoided if the right platform is chosen at the outset.

 

Q: What are some of the popular platforms for developing natural language conversational systems that you will evaluate?

A: Popular platforms include Microsoft LUIS, Nuance Mix, the Alexa Skills Kit, Google Dialogflow, IBM Watson Assistant, and Rasa.

 

Q: What are some of the criteria used in your evaluation?

A: Some criteria include: support for complex entities; ease of implementing open (mixed-initiative) dialog patterns; ability to measure and tune the NLU accuracy; ability to import/export annotated data; ease of annotation; number of built-in entities; and multilingual support.

For real, non-toy, applications, we need to look at cost, support for multiple developers, as well as support for the whole project life-cycle. That includes debugging capabilities, ability to run batch dialog tests, support for archiving and versioning applications, and what analytics are provided. Nice to haves are support for backend integration, including databases and web service interfaces.

 

Q: What are key learnings that attendees will take away from your presentation

A: The attendees will learn about the platform options for their applications and how to match platforms to their requirements.

 

Register for SpeechTEK today to learn more about The Engineering of Emotion. 

Page1 of 1
Learn more about the companies mentioned in this article in the Speech Technology Buyer's Guide:
Learn more about the companies mentioned in this article in the Vertical Markets Guide: