Tag: statistics

TEDxNicosia 2013 is just a few short hours away.   As I mentioned previously, I am very excited, and I keep thinking about it.  One particular thought was bugging me all day today – how are the speakers being selected, and is there anything common among them? Do they share any specific knowledge or experience, or personal characteristics?  Not knowing any of the speakers personally, I decided to go for some fun PHP scripting rather than any serious research.  It’s Friday after all!

The result is this little project.  I basically took the 12 speaker profiles directly from the TEDxNicosia speakers page, and used it as my source data.  Each profile is saved into a text file with the name of the speaker.  Then I ran some simple analysis on those text files.   First, I wanted to see if their profile texts were sharing any common words.  That would be an indication, right?  Obviously, I had to filter out some words like ‘as’, ‘and’, and ‘he’ (see a full list of filtered out words).  For the rest, here is the top 20 most common words (by the way, the script reports the names of speakers as well, but I took it out for clarity and simplicity):

1. Cyprus, shared by 11 out of 12 profiles;
2. years, shared by 10 / 12;
3. university, shared by 8 / 12;
4. international, shared by 7 / 12;
5. world, shared by 7 / 12;
6. work, shared by 6 / 12;
7. national, shared by 5 / 12;
8. media, shared by 5 / 12;
9. currently, shared by 5 / 12;
10. including, shared by 5 / 12;
11. well, shared by 5 / 12;
12. all, shared by 5 / 12;
13. life, shared by 5 / 12;
14. first, shared by 5 / 12;
15. people, shared by 5 / 12;
16. USA, shared by 5 / 12;
17. development, shared by 4 / 12;
18. London, shared by 4 / 12;
19. business, shared by 4 / 12;
20. experience, shared by 4 / 12;

Interesting, isn’t it?  The easiest to notice for me is geography.  The most shared word is Cyprus, which is not surprising, because the TEDxNicosia event is happening, here, in Cyprus, and because most of the speakers either live here, or were born here, or moved here.  the other two geographical highlights are the USA and UK (London specifically).  These are the most influential, however there are indications of other travel (national, international, world).

One other thing which stands out is hard work.  It is suggested by work, all, life, development, business, and experience.  It sounds like all these people know what they are talking about.  Especially if you throw in university in there.  Also, first is indicative of either trying new things or of leading somewhere.

The rest might also mean something, but they don’t stand out so much.  At least not to me.   Except maybe if I put together media and people.  Then there is a sort of social suggestion.

After reading speakers’ profiles, I think the above is pretty accurate.  Even if it wasn’t, accuracy wasn’t exactly the point.  The whole thing is more of technical entertainment piece.  Oh, by the way, that reminds me.  What does TED stand for?  Technology, Entertainment, Design.  While we are looking at speaker profile words, why don’t we try and see if the TED words are in there too.  A bit more of coding, and here is what I get:

• technology is represented by 3 out of 12 speakers;
• entertainment is not represented by anyone;
• design is represented by 2 out of 12;

Doesn’t sound too good?  Well, that’s because these numbers have very little to do with the actual speakers.  The source data were speaker profiles, which are only a few words long.  If these were worded even slightly different, the results would be completely different.  Just to give you an indication – even though the word ‘entertainment’ haven’t been used, a few other words, such as ‘music’, ‘dance’, ‘film’, ‘book’ were used plenty, and these can easily be used near entertainment.

Now that Friday night is quickly turning into Saturday morning, I think I should grab a few hours of sleep and drive out to Nicosia.  See you all there, or see you all after!