Thursday, August 11, 2016

Chatbot info

Search for "chatbot machine learning"

-----------------------------

ChatterBot 0.4.5

An open-source chat bot program written in Python.

ChatterBot is a machine learning, conversational dialog engine.

https://github.com/gunthercox/ChatterBot
https://pypi.python.org/pypi/ChatterBot

This is the documentation. Read this:
http://chatterbot.readthedocs.io/

ChatterBot is a machine-learning based conversational dialog engine build in Python which makes it possible to generate responses based on collections of known conversations. The language independent design of ChatterBot allows it to be trained to speak any language.

Warning
The JsonDatabaseAdapter is not intended for use with large amounts of data. You may expirience serious performance problems if the size of this database becomes too large.




-----------------------------------

CNN bot at FB Messenger.
Kind of interesting. Enter a word and it shows news.

https://www.messenger.com/t/CNN/

Also, Tech Crunch messenger:
https://www.messenger.com/t/techcrunch/

---------------------------------

http://www.estherbot.com/

 Chatbot that tells Esther's life story, bio, resume.
But kind of hard to use.

---------------------------------

The Complete Beginner’s Guide To Chatbots

Everything you need to know.

https://chatbotsmagazine.com/the-complete-beginner-s-guide-to-chatbots-8280b7b906ca#.ipaapc3aj 

Lengthy explanation about how chatbot (messaging) is important for businesses. 

Great services you can use to build your bot:
  • wit.ai (bought by Facebook)
  • Wit is free, including for commercial use. So both private and public Wit apps are free and are governed by our terms.
  • Supports Korean and many other languages.

  • Do you have the ownership on my data?

    You have the ownership of your data but you agree to let us use them to improve Wit. If your app is private, your data (intents, entities, expressions, logs) will be accessible only by you and the developers you decide to share your app with. If your app is open, you agree to share some data.  
  • Seems like a lot of manual work needed?
  • howdy’s botkit (raised $1.5+ mil in funding)
  •  
  • api.ai (raised $8.6+ mil in funding)
  • The leader in conversational voice interfaces for mobile, web and devices.
  •  
  •  
  • textit.in
  • With Flows, anybody can create engaging SMS and voice applications without the need of a programmer or expensive consulting company.  

  • Motion.ai
  • Free for up to 1000 msgs/mo
  •  
  • Chatfuel (Ycombinator company)
  • I used this to create a bot at Facebook. See ss.
  • Chatfuel’s bot-building platform is free
    If you plan on hitting more than 100,000 conversations a month, please talk to us at premium@chatfuel.com.
    Used by UBER, several pro sports teams
     
Other Resources:

 ------------------------------------

Deep Learning for Chatbots, Part 1 – Introduction

http://www.wildml.com/2016/04/deep-learning-for-chatbots-part-1-introduction/

Skimmed this. Intro stuff.

Deep Learning for Chatbots, Part 2 – Implementing a Retrieval-Based Model in Tensorflow

But why would you want to build a retrieval-based model if you can build a generative model? Generative models seem more flexible because they don’t need this repository of predefined responses, right?
The problem is that generative models don’t work well in practice. At least not yet. Because they have so much freedom in how they can respond, generative models tend to make grammatical mistakes and produce irrelevant, generic or inconsistent responses. They also need huge amounts of training data and are hard to optimize. The vast majority of production systems today are retrieval-based, or a combination of retrieval-based and generative. Google’s Smart Reply is a good example. Generative models are an active area of research, but we’re not quite there yet. If you want to build a conversational agent today your best bet is most likely a retrieval-based model.

The Ubuntu Dialog Corpus

In this post we’ll work with the Ubuntu Dialog Corpus (paper, github). The Ubuntu Dialog Corpus (UDC) is one of the largest public dialog datasets available. It’s based on chat logs from the Ubuntu channels on a public IRC network. The paper goes into detail on how exactly the corpus was created, so I won’t repeat that here. However, it’s important to understand what kind of data we’re working with, so let’s do some exploration first.
The training data consists of 1,000,000 examples, 50% positive (label 1) and 50% negative (label 0). Each example consists of a context, the conversation up to this point, and an utterance, a response to the context. A positive label means that an utterance was an actual response to a context, and a negative label means that the utterance wasn’t – it was picked randomly from somewhere in the corpus.
-------------------------------------

From Quora:

Marcus L Endicott, Robopsychologist @ Meta-Guide.com
34.9k Views · Most Viewed Writer in Artificial Intelligence with 330+ answers

The AI Zone forum at Chatbots.org is a happening place.

There are Quora topics for:


See also my own webpage:

Updated Jul 28 
----------------------------------------


 Coursera

Build Intelligent Applications
Master machine learning fundamentals in five hands-on courses.


https://www.coursera.org/specializations/machine-learning?utm_source=gg&utm_medium=sem&utm_campaign=machine_learning_search_intl&campaignid=426787857&device=c&keyword=online%20course%20machine%20learning&matchtype=b&network=g&devicemodel=&adpostion=1t2&hide_mobile_promo&gclid=Cj0KEQjwxLC9BRDb1dP8o7Op68IBEiQAwWggQIYZ0ix9RD4xhPK0B9xe4I_vk5a4ufITLr7uSESFIDEaAjom8P8HAQ
--------------------------------------
http://venturebeat.com/2016/07/20/donald-trump-hillary-clinton-chatbot-sapientx/

This Donald Trump chatbot is great… really, really great. It’s unbelievable.

SapientX, a company that has been in stealth mode for the last year, has created chatbots for Donald Trump and Hillary Clinton that provide words directly from the candidate’s mouth on topics ranging from abortion to taxes and terrorism.

The chatbots are able to answer questions in text or voice about roughly 100 topics, like: “What do you think about the Black Lives Matter movement?” or “Do you think that women should be paid as much as men?” A full list of sample questions and topics is available here.

The two bots can be found on AskHillaryandDonald.com, and they draw on audio clips taken from public media since Clinton and Trump declared their candidacy for the presidency more than a year ago, said Jonathan Hirshon, a member of SapientX’s board of advisors.

---------------------------------------------------------------------------

 

 




Posted by at

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.

Subscribe to: Post Comments (Atom)