<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: Chatbot Architecture</title>
	<atom:link href="http://pavel.surmenok.com/2016/09/11/chatbot-architecture/feed/" rel="self" type="application/rss+xml" />
	<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/</link>
	<description></description>
	<lastBuildDate>Sat, 13 Apr 2019 17:14:00 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.5.1</generator>
	<item>
		<title>By: surmenok</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-168</link>
		<dc:creator>surmenok</dc:creator>
		<pubDate>Fri, 21 Sep 2018 05:07:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-168</guid>
		<description><![CDATA[There was research showing that using word+POS tag works better than just a word for text classification tasks.]]></description>
		<content:encoded><![CDATA[<p>There was research showing that using word+POS tag works better than just a word for text classification tasks.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Roma Jain</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-167</link>
		<dc:creator>Roma Jain</dc:creator>
		<pubDate>Wed, 19 Sep 2018 02:58:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-167</guid>
		<description><![CDATA[Awesome article! 
Should the words tagged as entities be used instead of words for intent classification?]]></description>
		<content:encoded><![CDATA[<p>Awesome article!<br />
Should the words tagged as entities be used instead of words for intent classification?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Giovanni Morrone</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-136</link>
		<dc:creator>Giovanni Morrone</dc:creator>
		<pubDate>Thu, 13 Jul 2017 16:33:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-136</guid>
		<description><![CDATA[Hi vvvlad, I understand. Thanks for your reply. Good luck for the continuation of the project!]]></description>
		<content:encoded><![CDATA[<p>Hi vvvlad, I understand. Thanks for your reply. Good luck for the continuation of the project!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: vvvlad</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-132</link>
		<dc:creator>vvvlad</dc:creator>
		<pubDate>Fri, 30 Jun 2017 11:00:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-132</guid>
		<description><![CDATA[hey,
this is an internal project of which I can&#039;t provide to many details.
The data can be anything, from a simple knowledge article to a FAQ items with numerous relevant questions pointing to a single answer. The data sources are managed by other teams.
The main goal of the project is to create a code that would generate CS to reduce maintanance of data in the future.
I look at each answer as an intent, and the questions or keywords or details are the sources for ngram extraction.]]></description>
		<content:encoded><![CDATA[<p>hey,<br />
this is an internal project of which I can&#8217;t provide to many details.<br />
The data can be anything, from a simple knowledge article to a FAQ items with numerous relevant questions pointing to a single answer. The data sources are managed by other teams.<br />
The main goal of the project is to create a code that would generate CS to reduce maintanance of data in the future.<br />
I look at each answer as an intent, and the questions or keywords or details are the sources for ngram extraction.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Giovanni Morrone</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-131</link>
		<dc:creator>Giovanni Morrone</dc:creator>
		<pubDate>Wed, 28 Jun 2017 15:38:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-131</guid>
		<description><![CDATA[I don&#039;t understand what exactly you are doing. What is the structure of sources? How do you generate n-gram conditions and their associate replies? If exists a repo of your project I could understand better.]]></description>
		<content:encoded><![CDATA[<p>I don&#8217;t understand what exactly you are doing. What is the structure of sources? How do you generate n-gram conditions and their associate replies? If exists a repo of your project I could understand better.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: vvvlad</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-130</link>
		<dc:creator>vvvlad</dc:creator>
		<pubDate>Wed, 28 Jun 2017 13:38:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-130</guid>
		<description><![CDATA[Not sure what details you are looking for.
Most of my work until now was concentrated on python data preparation and less with CS, but in parallel I test a few things with CS as well.
My current plan is to generate many (&lt;&gt;) conditions (which can result in multiple intents) and add them to list, where I would get the top appearances of intents and reply with the most relevant intent result.
The rules are generated automatically, so they can&#039;t be specific. This is a bit like decision tree, but not exactly.]]></description>
		<content:encoded><![CDATA[<p>Not sure what details you are looking for.<br />
Most of my work until now was concentrated on python data preparation and less with CS, but in parallel I test a few things with CS as well.<br />
My current plan is to generate many (&lt;&gt;) conditions (which can result in multiple intents) and add them to list, where I would get the top appearances of intents and reply with the most relevant intent result.<br />
The rules are generated automatically, so they can&#8217;t be specific. This is a bit like decision tree, but not exactly.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Giovanni Morrone</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-129</link>
		<dc:creator>Giovanni Morrone</dc:creator>
		<pubDate>Wed, 28 Jun 2017 11:03:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-129</guid>
		<description><![CDATA[I&#039;m studying ChatScript and your work seems very interesting. Could you tell me more details?]]></description>
		<content:encoded><![CDATA[<p>I&#8217;m studying ChatScript and your work seems very interesting. Could you tell me more details?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: surmenok</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-127</link>
		<dc:creator>surmenok</dc:creator>
		<pubDate>Tue, 23 May 2017 20:04:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-127</guid>
		<description><![CDATA[Sounds interesting. Please let me know how it works in the real world.]]></description>
		<content:encoded><![CDATA[<p>Sounds interesting. Please let me know how it works in the real world.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: vvvlad</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-126</link>
		<dc:creator>vvvlad</dc:creator>
		<pubDate>Tue, 23 May 2017 08:07:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-126</guid>
		<description><![CDATA[I&#039;m trying to do something a bit similar.
My plan is to create a python code which would generate chatscript script by pulling and prepossessing the knowledge information from different sources. As a first step I generated simple bi and tri-grams, and it looks very promising.]]></description>
		<content:encoded><![CDATA[<p>I&#8217;m trying to do something a bit similar.<br />
My plan is to create a python code which would generate chatscript script by pulling and prepossessing the knowledge information from different sources. As a first step I generated simple bi and tri-grams, and it looks very promising.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: surmenok</title>
		<link>http://pavel.surmenok.com/2016/09/11/chatbot-architecture/#comment-112</link>
		<dc:creator>surmenok</dc:creator>
		<pubDate>Tue, 21 Feb 2017 04:02:00 +0000</pubDate>
		<guid isPermaLink="false">http://pavel.surmenok.com/?p=196#comment-112</guid>
		<description><![CDATA[I wouldn&#039;t recommend ChatScript for a knowledge base system. Typically chatbot developers use search technology for this, for example Elastic Search. What technology do you currently use for this?]]></description>
		<content:encoded><![CDATA[<p>I wouldn&#8217;t recommend ChatScript for a knowledge base system. Typically chatbot developers use search technology for this, for example Elastic Search. What technology do you currently use for this?</p>
]]></content:encoded>
	</item>
</channel>
</rss>
