Data

...now browsing by category

 

“Oracle of Omaha” — A Must Read

Saturday, June 30th, 2007

In a world of information overload there are few must reads, but if you appreciate what Warren Buffett has done then here is some really good stuff by Jeff Matthews

Jeff has a series of posts entitled “Pilgrimage to Omaha.” I only read two posts in a series of 10 (plus 2 extras for intro and conclusion) and I am super excited to read the rest (that rarely happens)— here is but one of the many laconic and insightful points Jeff makes:

“Thus it is that whether he’s talking about food companies or airlines or newspapers or oil companies, Buffet has clearly made it his business to identify the single most important variable for each business—and knowing those crucial variables, he can determine whether the values offered in the stock market at any given time are attractive, or not. Without using a spreadsheet.”

A Quick Comparison of Google and Indeed.com Trends

Saturday, February 3rd, 2007

Steve Rubel has a good post on the use of basic (yet powerful) analytic tools such as Google Trends and Indeed to develop insight into whether or not certain technologies are hype or here to stay.  While there is no doubt that these tools are powerful I have some concerns with the use of the Indeed data for a few reasons:

  • Unlike Google Trends or Flickr Camera Finder which rely on data that is directly provided by users on their respective sites the data on which Indeed trends is based on is based on their crawled data and therefore very suspect (full disclosure I did work in the Yahoo!’s HotJobs group).  Lets go one level deeper and look at the Indeed corpus and why it is not suitable for analytics and trending:
    • Hiring managers and recruiters flood their job descriptions with buzzwords in order to rank higher on Careerbuilder, Monster, HotJobs, etc. hence generating hiring trends based on the occurrence of certain keywords in a job description won’t tell you that much and for certain keywords will likely lead you in the wrong direction.
    • Aggregating job listings via a crawl (which is what Indeed does) is not easy and it yields tens of duplicates per job (every recruiter posts their jobs to multiple sites).

By the way Indeed is a pretty good job search engine, this trends feature is a bit out of place.  It would be far more interesting if they provided trends on what their users were searching for.

Coffee Consumption by Country

Wednesday, January 24th, 2007

Interesting data courtesy of IBM’s new Many Eyes project. 

  • I would have thought that the South American countries such as Brazil and Argentina would make the top 10.
  • Would be interesting to see the per capita consumption of coffee — I would guess France and Italy would lead the way.  Then again with our Starbucks addiction…
  • By the way if you are looking for some great coffee —not the Starbucks crap — then check out Peets