<?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: Good Interaction Design II: Just Ask</title>
	<atom:link href="http://irgupf.com/2009/11/04/good-interaction-ii-just-ask/feed/" rel="self" type="application/rss+xml" />
	<link>http://irgupf.com/2009/11/04/good-interaction-ii-just-ask/</link>
	<description>Information Retrieval Research, Issues, and Discussion</description>
	<lastBuildDate>Tue, 14 Jun 2011 14:26:42 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.1.3</generator>
	<item>
		<title>By: Information Retrieval Gupf &#187; Kasparov and Good Interaction Design</title>
		<link>http://irgupf.com/2009/11/04/good-interaction-ii-just-ask/comment-page-1/#comment-7143</link>
		<dc:creator>Information Retrieval Gupf &#187; Kasparov and Good Interaction Design</dc:creator>
		<pubDate>Mon, 25 Jan 2010 22:59:20 +0000</pubDate>
		<guid isPermaLink="false">http://irgupf.com/?p=1032#comment-7143</guid>
		<description>[...] similar to some of the other results I&#8217;ve reported on in the past.  For example, see this paper by Amatriain: Data is always important, but what struck me in the writeup was his discovery that the biggest [...]</description>
		<content:encoded><![CDATA[<p>[...] similar to some of the other results I&#8217;ve reported on in the past.  For example, see this paper by Amatriain: Data is always important, but what struck me in the writeup was his discovery that the biggest [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: vaultis.com &#187; Blog Archive &#187; 7 Days of Search and Social; the ultimate update</title>
		<link>http://irgupf.com/2009/11/04/good-interaction-ii-just-ask/comment-page-1/#comment-5490</link>
		<dc:creator>vaultis.com &#187; Blog Archive &#187; 7 Days of Search and Social; the ultimate update</dc:creator>
		<pubDate>Wed, 11 Nov 2009 14:45:33 +0000</pubDate>
		<guid isPermaLink="false">http://irgupf.com/?p=1032#comment-5490</guid>
		<description>[...] Good Interaction Design II: Just Ask  - IR Gupf [...]</description>
		<content:encoded><![CDATA[<p>[...] Good Interaction Design II: Just Ask  - IR Gupf [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jeremy</title>
		<link>http://irgupf.com/2009/11/04/good-interaction-ii-just-ask/comment-page-1/#comment-5428</link>
		<dc:creator>jeremy</dc:creator>
		<pubDate>Thu, 05 Nov 2009 00:57:43 +0000</pubDate>
		<guid isPermaLink="false">http://irgupf.com/?p=1032#comment-5428</guid>
		<description>Bob: Good points.  You&#039;re saying, if I understand correctly, that if you phrase how you elicit the question, you can actually make the job of the algorithm much easier?</description>
		<content:encoded><![CDATA[<p>Bob: Good points.  You&#8217;re saying, if I understand correctly, that if you phrase how you elicit the question, you can actually make the job of the algorithm much easier?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Bob Carpenter</title>
		<link>http://irgupf.com/2009/11/04/good-interaction-ii-just-ask/comment-page-1/#comment-5424</link>
		<dc:creator>Bob Carpenter</dc:creator>
		<pubDate>Wed, 04 Nov 2009 19:25:39 +0000</pubDate>
		<guid isPermaLink="false">http://irgupf.com/?p=1032#comment-5424</guid>
		<description>This is actually why I went to work for SpeechWorks.  I&#039;d just seen Mike Phillips, CTO and co-founder, give a talk at ASRU emphasizing the fact that spoken dialogue systems present mainly an HCI problem, not an electrical engineering problem.  It was a breath of fresh air compared to EE-heavy Bell Labs research (being in the Rockies rather than NYC may also have helped).

But we have to be careful.  The single most highly correlated feature with task completion rates is speech recognition accuracy.  You might think that&#039;d motivate a garage full of electrical engineers toiling away on accuracy.  Well, it does.  But what it also motivates is a team of user interface designers toiling away on dialogue strategies to elicit understandable response from users and present data to them in an understandable way.  And to develop dialogue strategies to deal with the inevitable errors.

For instance, asking a zip code before an address is a huge help.  Zip codes are five digits, which are relatively easy to recognize.  Then states, then cities.  The state fact-checks the zip, then together restrict the possible cities.  The research problem is to just say your address and transcribe it.  This problem hasn&#039;t been solved by better speech recognition.

Built-in design also helps.  Credit cards have checksums, which can be used to aid recognition.  Who knows how many errors this has saved, and not just from speech reco.</description>
		<content:encoded><![CDATA[<p>This is actually why I went to work for SpeechWorks.  I&#8217;d just seen Mike Phillips, CTO and co-founder, give a talk at ASRU emphasizing the fact that spoken dialogue systems present mainly an HCI problem, not an electrical engineering problem.  It was a breath of fresh air compared to EE-heavy Bell Labs research (being in the Rockies rather than NYC may also have helped).</p>
<p>But we have to be careful.  The single most highly correlated feature with task completion rates is speech recognition accuracy.  You might think that&#8217;d motivate a garage full of electrical engineers toiling away on accuracy.  Well, it does.  But what it also motivates is a team of user interface designers toiling away on dialogue strategies to elicit understandable response from users and present data to them in an understandable way.  And to develop dialogue strategies to deal with the inevitable errors.</p>
<p>For instance, asking a zip code before an address is a huge help.  Zip codes are five digits, which are relatively easy to recognize.  Then states, then cities.  The state fact-checks the zip, then together restrict the possible cities.  The research problem is to just say your address and transcribe it.  This problem hasn&#8217;t been solved by better speech recognition.</p>
<p>Built-in design also helps.  Credit cards have checksums, which can be used to aid recognition.  Who knows how many errors this has saved, and not just from speech reco.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

