Chatbots Should Be An Abstraction Of Human Conversation

10 Elements Of A Conceptual Process To Derive General Conversational Rules & Concepts

Introduction

Image for post
Image for post
Still life by Dutch painter, Henk Helmantel. I found viewing his paintings in real life jarring. You know it is an abstraction of reality, yet appearing to be real.

1️⃣ Digression

Image for post
Image for post
Digression Example Where a User Momentarily Digress

2️⃣ Disambiguation

Image for post
Image for post
IBM Watson Assistant Example of Disambiguation Between Dialog Nodes
Image for post
Image for post
IBM Watson Assistant Initial Configuration for Disambiguation

3️⃣ Auto Learning

Image for post
Image for post
Practical Example Of How The Menu Is Be Presented

4️⃣ Domain & Irrelevance

Negate False Intent Assignment

OOD & ID

Image for post
Image for post
An example of the IBM Watson Assistant testing and training interface. User utterances can be assigned to an existing intent, or marked as irrelevant.

5️⃣ Compound Intents

Image for post
Image for post
More Complex Entity Detection with Linked Entities

6️⃣ Anthropomorphize

IBM Virtual Voice Creator

7️⃣ Named Entities

Image for post
Image for post
Extracting Named Entities from a Sentence
Image for post
Image for post
Detail On Each Named Entity Detected

8️⃣ Mixed Modality & Conversational Components

Image for post
Image for post
Facebook Messenger with Feature Rich Conversational Components
Image for post
Image for post
WhatsApp With Far Less Conversational Components

9️⃣ Contextual Entities

Image for post
Image for post
Annotate Entities In IBM Watson Assistant

🔟 Variation

Conclusion

Written by

NLP/NLU, Chatbots, Voice, Conversational UI/UX, CX Designer, Developer, Ubiquitous User Interfaces. www.cobusgreyling.me

Get the Medium app