<?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: Product Review: Using Brandwatch for Social Media Monitoring</title>
	<atom:link href="http://www.monitoring-social-media.com/product-review-using-brandwatch-for-social-media-monitoring/feed" rel="self" type="application/rss+xml" />
	<link>http://www.monitoring-social-media.com/product-review-using-brandwatch-for-social-media-monitoring</link>
	<description>The Latest News &#38; Reviews on Social Media Monitoring and Measurement, Influence Analysis, Sentiment Detection, Data Analysis, Monitoring Tools, Reputation Management and PR Measurement.</description>
	<lastBuildDate>Thu, 01 Mar 2012 17:39:00 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.3</generator>
	<item>
		<title>By: Aisha Arun</title>
		<link>http://www.monitoring-social-media.com/product-review-using-brandwatch-for-social-media-monitoring#comment-93</link>
		<dc:creator>Aisha Arun</dc:creator>
		<pubDate>Tue, 20 Dec 2011 15:34:00 +0000</pubDate>
		<guid isPermaLink="false">http://www.monitoring-social-media.com/?p=313#comment-93</guid>
		<description>We have been using the Enterprise version of Brandwatch and have been happy about the overall performance of the tool so far. As pointed out, the sentiment is sometimes way off, but I believe there is room for improvement there on the technology front itself for everyone in this field.

Agree with the views in this review.</description>
		<content:encoded><![CDATA[<p>We have been using the Enterprise version of Brandwatch and have been happy about the overall performance of the tool so far. As pointed out, the sentiment is sometimes way off, but I believe there is room for improvement there on the technology front itself for everyone in this field.</p>
<p>Agree with the views in this review.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Luke Brynley-Jones</title>
		<link>http://www.monitoring-social-media.com/product-review-using-brandwatch-for-social-media-monitoring#comment-92</link>
		<dc:creator>Luke Brynley-Jones</dc:creator>
		<pubDate>Tue, 20 Dec 2011 15:09:00 +0000</pubDate>
		<guid isPermaLink="false">http://www.monitoring-social-media.com/?p=313#comment-92</guid>
		<description>Hi Joel - Understood and agreed. I think our point here is that nobody has cracked sentiment yet, so we&#039;ll be hard pushed to give anyone an A-star rating on that front. Enabling customers to &quot;teach&quot; their systems is far better than a Black Box approach though, so you&#039;ll certainly be ahead on points.</description>
		<content:encoded><![CDATA[<p>Hi Joel &#8211; Understood and agreed. I think our point here is that nobody has cracked sentiment yet, so we&#8217;ll be hard pushed to give anyone an A-star rating on that front. Enabling customers to &#8220;teach&#8221; their systems is far better than a Black Box approach though, so you&#8217;ll certainly be ahead on points.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Joel Windels</title>
		<link>http://www.monitoring-social-media.com/product-review-using-brandwatch-for-social-media-monitoring#comment-91</link>
		<dc:creator>Joel Windels</dc:creator>
		<pubDate>Tue, 20 Dec 2011 15:05:00 +0000</pubDate>
		<guid isPermaLink="false">http://www.monitoring-social-media.com/?p=313#comment-91</guid>
		<description>We do indeed recommend using manual mark-up to improve the accuracy of the sentiment analysis. Although we work hard on it, and have been pioneering a number of novel approaches to automated sentiment analysis (using both rules-based classifiers and machine-learning), we are still only accurate around 60-70% of the time. 

This, however, is common in the industry, and there are no tools claiming to have accuracy above that. It can be useful to paint a general picture of sentiment in data sets - even if individual mentions are occasionally wrong - when identifying trends and looking for insights.

Thanks for featuring us,

Joel
Community Manager at Brandwatch</description>
		<content:encoded><![CDATA[<p>We do indeed recommend using manual mark-up to improve the accuracy of the sentiment analysis. Although we work hard on it, and have been pioneering a number of novel approaches to automated sentiment analysis (using both rules-based classifiers and machine-learning), we are still only accurate around 60-70% of the time. </p>
<p>This, however, is common in the industry, and there are no tools claiming to have accuracy above that. It can be useful to paint a general picture of sentiment in data sets &#8211; even if individual mentions are occasionally wrong &#8211; when identifying trends and looking for insights.</p>
<p>Thanks for featuring us,</p>
<p>Joel<br />
Community Manager at Brandwatch</p>
]]></content:encoded>
	</item>
</channel>
</rss>

