chatbots and voice conversational interfaces with amazon alexa, neo4j and graphaware nlp

Post on 22-Jan-2018

1.827 Views

Category:

Technology

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

Hello humans !

Forget everything !

Everything you know…

Everything you learned…

is OVER !!!

Welcome to a brand new world…

The world of VOICE

From Alexa

To Google Home,

In-car systems,

and smartwatches

Touch is no longer the primary UI

The touch era is OVER !

Chatbot and Conversational Experiences with Amazon

Alexa, Neo4j and GraphAware NLP

GraphDB Meetup Czech RepublicJune 14th, 2017

Christophe Willemsen - @ikwattro

Program of tonight

• Chatbots history

• Voice-driven chatbot anatomy

• Let’s build our conference assistant

• Introduction to Natural Language Processing

• Let’s chat..bot

Chatbots history

• Alan Turing Test

Chatbots history

• Eliza (1966) mimicked human conversations by matching user

prompts to scripted responses. It was able, at least for a time, to

pass the Turing test

Chatbots history

• Parry (1972) – Inexplicably simulated a person with paranoid

schyzophrenia. Parry was more advanced than Eliza.

Chatbots history

• 1988 – Jabberwacky

• 1992 - Dr. Sbaitso

• 1995 – A.L.I.C.E

• 2001 – Smarterchild

• 2006 – IBM’s Watson

• 2010 – SIRI

• 2012 – Google NOW

• 2016 – Bots for Messenger, TAY

• And Alexa ? Ask her !

Chatbot anatomy

• Voice Commands

• Intent detection

• Intent handling

• Response

Our conference assistant

• Tell me how many sessions ?

• Tell me how many sessions about a specific topic ?

• Recommend me a session based on my topic preferences

• Find me a session about a specific topic with NLP

Alexa should be able to :

Dataset usedGraphconnect London 2017 schedule

Voice command

• Give a name to our Alexa skill -> CONFERENCE ASSISTANT

• Define our first intent name (code) -> talksCount

• Define our first utterance

ALEXA, ASK CONFERENCE ASSISTANT

(skill invocation)

HOW MANY SESSIONS TODAY ?

(skill utterance)

INTENT DETECTION

SPEECH TO TEXT

TEXT TO SPEECH

LAMBDA FUNCTIONOR

YOUR API SERVER

ALEXA VOICE SERVICE

Utterances

• How many sessions ?

• How much sessions ?

• How many talks at the conference ?

• How many talks today ?

Intent detection

• 2 main types of detection :

-> Pattern matching (how | count)* many (session | talks) today?

-> Classification machine learning, neural networks, word2vec, ..

Alexa send to the skill API

Your skill returns a text response

Intent detection

• 2 main types of detection :

-> Pattern matching (how | count)* many (session | talks) today?

-> Classification machine learning, neural networks, word2vec, ..

First intent test

ALEXA, HOW MANY SESSIONS TODAY ?

ALEXA, ASK CONFERENCE ASSISTANT

(skill invocation)

HOW MANY SESSIONS TODAY ?

(skill utterance)

INTENT DETECTION

SPEECH TO TEXT

TEXT TO SPEECH

LAMBDA FUNCTIONOR

YOUR API SERVER

ALEXA VOICE SERVICE

Slots

• Variables in the utterances

• Should be filled with possible values

• Slots are sent in the intent request to your API

Slots

HOW MANY SESSIONS ABOUT {topic}

Second intent test

ALEXA, HOW MANY SESSIONS ABOUT

PERFORMANCE ?

Adding personalization

• If you have informations about the attendees and their topic

preferences, you can easily get the sessions in accordance with it.

Adding personalization

Adding personalization

Third intent test

ALEXA, RECOMMEND ME A SESSION

GraphAware NLP

• Graph based Natural Language Processing framework

• Information Extraction, Sentiment Analysis, Data enrichment

(ontologies and concepts), similiraty computation, knowledge

enrichment, …

• Beta program, first public open-source version to be released very

soon

Enriching with concepts

Enriching with concepts

Tackling the last intent

• Question / response game

• Treat question as text and process NLP

• Relate question text with Session abstracts texts

• Special type of utterance with only one slot {text}

Demo

Fun / Ideas with Alexa

Resources

• https://github.com/alexylem/jarvis

Thank you !

Questions?

graphaware.com

@graph_aware

top related