So many blogs to read, so little time. So how do you decide which ones to read?
In two words: blog summaries.
Here's the problem. If you're anything like me you may have a ton of RSS feeds in your RSS reader. Perhaps you have the posts of a few choice blogs going straight to your inbox. Either way, over time the RSS feeds start to stack up and you quickly become inundated with unread items. As I write this Reeder tells me I have 718 unread items.
That's just a few 😉
A number of years ago I built an app to try to handle more than 300 blogs I followed. The app did the following:
- Automatically pulled in RSS feeds
- Went to the web pages from the feeds and scraped the content
- Cleaned up the content
- Put it all into a custom-built search engine (thank you Nutch)
- Ran analysis on the content comparing it to other posts I had stored to see if more than one person was talking about the same topic
Needless to say it was a jumble that worked better than expected but not well enough.
I have the same problem today, namely a ton of blog feeds I want to follow but not nearly enough time to read all the posts. And frankly, a title doesn't always tell you if the post is worth your reading time.
So what's a busy entrepreneur to do? Code an app that summarized blog posts using the latest in natural language processing and text analytics. In short, automating the summarization of blog posts.
Here's an example.
Andy Sontag wrote an excellent post on the FearLess Revolution blog about the architecture of authenticity. I recommend you read the entire thing, however here is the summary my app gave me:
I have observed friends coming together around the purpose of creating a truly meaningful experiences for others. Giving an experience to bring people together was deeply inspiring. It is not the opening of an extravagant gift from my parents, or biting into my mothers famous holiday bread, or watching the trivially small ball drop at the Yellow Springs new years celebration that I cherish so deeply; rather it is sharing these experiences with others that makes the experience important. Thus the best gift I could think of for my friends and family – was a gift that everyone would share and enjoy together – build social bonds and happiness. I invited 40 of my parent’s and brother’s closest friends. Another friend warmed up apple cider for everyone to enjoy. The experience was special not only for my family, but also for everyone that was there. It is important that we realize our agency to create and shape these experiences for our friends and family.
Hot damn that's awesome! At least I think it is. Basically my app boiled down 649 words to 161. From this summary I can tell that I definitely want to read the entire post.
Here's another example.
Christopher Penn is the only marketer I know who combines World of Warcraft with marketing lessons. That endears me to the man. So let's take a look at one of his latest posts – What do Klout scores say about purchase intent? [NEW STUDY]. Here's the summary:
What do Klout scores say about purchase intent? I also asked for your Twitter handle so I could follow you back and do some data analysis. In that spirit, I took the respondents of the survey, 67 of them who provided Twitter handles, and ran Klout and PeerIndex on them with the simple question: does influence have any correlation to purchase intent? There’s a built-in bias to the survey – if you don’t like me, chances are you don’t follow me, and therefore wouldn’t have any purchase intent either. I draw the reasonable conclusion that influence scores are no predictor whatsoever of purchase intent.
In this case the app created a 101 word summary from a close to 500 word post. Now with this post the title tells you exactly what you can expect and is definitely intriguing enough to read. What I find uber interesting though is that my summarization algorithm (aka fancy math coding stuff) picked up on Chris' end conclusion.
So as they say in startup land here's the plan:
- Step 1: Build app
- Step 2: ??
- Step 3: Profit!
Ah that tricksy step 2. However step 2 in this case is make the entire process more automated so my app will pull in posts from my RSS feeds, create summaries, and then display them for me in a nice font that's easy on the eyes. Also improving the algorithm is high on the to do list too as list posts aren't handled very well.
After that the “profit” part comes in by turning it into a service.
Stay tuned for details on that by subscribing to blog updates via email below.