I’ve been a little curious about a company called Biz360 since I saw them listed on my feed subscriber list. Since the kids are happily playing next door with friends I decided to caste aside my cleaning supplies and check them out. Anything is better than housework huh?

Biz360 has a product called Market360 BlogView. Here’s a little bit of what it’s for:


BlogView is a measurement and market intelligence solution that enables you to keep track of what’s being said in the Blogosphere about your company, its products and competitors, and the trends that impact your business.

You can read the rest here

Sounds cool to me. If I had a company I would like to know what people think. A little bit of reliable information is good thing. I question the reliability of a blog post. But, it certainly wouldn’t hurt would it?

Well, unless it’s used to make important business decisions based on anonymous speculation. I guess it usually works out.

I have one main question. How accurate is it? Does the software used to index the posts know the difference between genuine sentiment and sarcasm? Many times, even humans have a hard time telling the difference.

For example, suppose I typed:

I stopped by Wal-Mart today and my daughter had to go potty. Don’t you just love how those Wal-Mart bathrooms smell? It was a pleasant experience. You know how much I love Wal-Mart.

Now, would that be counted as a +1 for WAL-MART? Would the WAL-MART people think I really like how their bathroom smells? My overactive brain can see commercials with happy people exiting the restrooms and big yellow dots bouncing around. Kinda scary isn’t it?
I think most people would realize that a reference to a public bathroom is usually negative, but does the software know that?

Ok, let’s try another:

The pricing on the Method cleaner is a little high, but that OK with me. I just have to clean less.

Would Method think that I really like it when they overprice a cleaner because I’m lazy and appreciate having a reason not to clean? Would they think I actually clean less because of the price? I clean less because I like my blog! And quite a few others.

I’m curious about this stuff. I’m curious about how human beings use their past experiences and social views to interpret writing. I am way more curious about how you can get a bunch of if-then statements and word associations to do it. I’m sure with a huge quantity of data you can make educated guesses but I don’t think you can really get in touch with a market this way. If the people at Biz360 can give me a little insight I would appreciate it. I would definitely give you a +1. Right now you are a neutral.

So how about it, do you think a computer can effectively analyze the content of a blog post and get a true reading on popular sentiment? Ok, that’s a little too nerdy. Feel free to answer anyway. I’m rather fond of nerds. How about, can you make a statement, dripping with sarcasm, which might confuse those bots indexing what we type? Come on, give it a try. Me likes well-written sarcasm. :)

4 Responses to “Drip a Little Sarcasm, Change the World”

  1. Thank you for the information. I have been reading through the marketiq blog. Interesting stuff. For some reason this fascinates me. I have to be careful or I’ll spend way to much time on this. When my kids get older I might be asking for a job!

    Golden Hacksaw huh? I have a hacksaw keychain from when I worked at [big ass company] Thanks again for the information!

  2. Thanks for the thoughtful question! You’re right–analyzing sentiment in blog posts is not easy. Besides sarcasm, challenges include slang, misspellings, and even simple negation.

    Our Point-of-View Sentiment engine works by using machine learning. This means that we manually identify examples of positive, negative, and neutral articles to help the engine formulate its automated ratings. The good news is that when many bloggers use sarcasm when referring to certain subjects, we incorporate that into the machine’s training. Our ratings are customized for every subject that our clients track, which means that our engine might rate blog posts about yogurt in a much different way than it rates blog posts about real estate.

    Overall, our sentiment engine is very accurate–enough that it garners high levels of satisfaction among our clients. Given the massive volume of blog content published every day, our ability to provide accurate aggregate sentiment ratings is one of our most important assets.

  3. Thank you for the link. I read through the reports and it was interesting. Makes me want to learn more about it. The BlogPulse site was awesome. I’m going to have to post about that. I spent way to much time playing around there. I also checked out your blog. Congratulations on the new baby!

  4. Hello,
    My name is Max Kalehoff and I work at a company called Nielsen BuzzMetrics, which does some things similar that Biz360 does. If you’re interested in learning more about blog monitoring, Forrester research recently published a big report on the various services (including Biz360 and us). You can download it for free here: http://www.nielsenbuzzmetrics......Report.pdf

    But you bring up some very good quetions.
    Cheers,
    Max Kalehoff

Leave a Reply

You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>