Word Cloud.... My new Language
Its really been a while since i updated this Blog as I was busy updating my other Blog on my cycling tour. As my previous post was on my trekking experience to Thadiyendamol, now its time for Science.
Very recently i came across a website called "tagcrowd.com" through my friend, and this site would show the variable font size of words relative to the frequency of occurrence of these words from a huge text file. When I used it for the 1st time by pasting a small article from google, I was not very much impressed with its logic as it seemed to be very simple. It goes through the entire text and calculates the frequency of occurrence for each word and displays the output by picking each word and displaying it with corresponding font size which is in proportion with the frequency. I just played for some time and completely ignored it (even forgot the name of the site).
After few days I came across a situation which was very adventitious. The latest movie of Manirathnam "Raavan" had mixed reviews both in Tamil and Hindi. Every site had their own view point and there were more than 100 sites. it was really a Hotch - Potch situation. I didn't have the patience or time to go through all the 100 sites, or even 10 sites with each review consisting of more than 500 words. Now the "tagcrowd.com" came as an abracadabra to my rescue. I simply tried to copy the articles from the 1st 20 sites and pasted it in tagcrowd and this was the output I got.
This output has a language in itself which doesn't make sense till we start analysing deeply. These are some of the observations that I made.
"Vikram" appears 9 times and Abhishek appears 7 times. So probably tamil version is better than Hindi as no review will keep stressing on an actor' name if it had a negative critic.
Raavan appears 4 times and Raavanan appears 5 times, not a big difference.
The word Superb appears thrice, so a good sign.
The words, "Tamil" and "Version" appears 4 times supporting my 1st statement.
The word "Watch" appears 6 times, again a good sign.
I prefer to Ignore the word "Beera" or "Veera" as its name of the protagonist and it has to appear in a every review.
"Dialogues" appear 5 times, again a positive sign.
"Watch" appears 6 times.
Water and Location has a frequency of 3, so probably the art and cinematography is good!!
Based on these observations, my hypothesis is the movie is defiantly watchable.
I was so excited and developed a deep affinity towards this new language. I chose a completely different topic for the 2nd time to confirm if what I just discovered is applicable to other set of articles. I will probably disclose the language post you do the analysis.
Yes, Its Deve Gowda. You see the word Yeddyurappa in such a big font for a reason. Not because he is the CM and Gowda was a former CM, but because recently Deve gowda used the word " Bloody Bastard" on Yeddyurappa. If i had increased the frequency to 100 from 50 words, it would have been visible.
And my last analysis was done on a very interesting personality, Mr Nithyananda, and formed a word cloud from 20 articles published.
Observation
The word with Maximum frequency is Nithyananda
The word with 2nd highest frequency s Gopika:)
We also have the word actress, ashram, CD and CID!!
The word suspect has a good frequency.
We have the key word "Video"
Hence, I need not go through all the 20 or 100 articles and I can just understand what the articles are taking about this man Mr.Nithyananda with just 50 words.
This new langugge really works. I have started using them to understand the performance of companies which will help me pick stocks. I will surely be a very frequent user of this tagcrowd .com. I am sure this will be of some help in future for you too as it was fructiferous for me. Now I strongly belive Steve Jobs quote "The dots will join at some point" So lets keep learning. All the best!!




I believe in god, and I need data for the rest. Keeping this in mind I wanted to know what makes him the “best batsmen” in the world that other top batsman in the world lack? And the best way I could relate his records with others was with data. The methodologies that I have chosen to study his performance is, by picking few of his outstanding records and comparing it with other few top batsmen in the world and analyze what stopped them from doing the same.

