Recently I was discussing about some topics related to corpus and search. After having the discussion i revised some of the points and found some interesting facts related to people preference and search. I am sharing some of the thoughts for you here.

Preference

Our discussion on corpus was related to a corpus in NLTK. It is a personal corpus collected from internet sources. My friend asked to get the dispersion plot of the words ‘handsome’, ‘rich’ and ‘sexy’ with NLTK. The result was quite disappointing. There was no ‘rich’ and ‘handsome’ found in the corpus but the word ‘sexy’ was there :-). After some re-examination we found that the words are there but occurred only once in the corpus and it occurred as uppercase. So why we got blank plot. When I pointed out, my friend told that ‘handsome’, ‘rich’ and ‘well-employed’ are the common specifications found in matrimony websites. It was a new information for me. Because I was wondering that how can we define ‘handsome’?? Again I googled for “rich well employed handsome groom” and found that all the ads which appearing contains these words :-) What a strange preference. If an automatic match maker is there how it will compare the request?? Because the views about ‘handsome’ will be different for different persons!!

Search

After making this experiment I discussed the matter with a colleague. He was telling something different on search patterns he found. He told that find the suggestions appearing for the search term ‘how to’ ! It was quite interesting the first suggestion was ‘how to get pregnant’ :-( . Even though much educative information is available in the internet what makes peoples to coin such a search phrase ?? Interestingly I found some other how to question ‘how to tell a girl that i love her’ ?? Huh Is it the result of internet addiction and isolation from society or friends. I collected the search result . The first hit was from wikihow.com . it contains guidelines,tips and tricks and warning :-)

Some of the how to queries are quite interesting. Observing such queries i wonder that whether we are properly using our information systems. Instead of getting educated still people are following some old myths and wrong information. Because people searches for ‘how to test virginity’ too.

Migrated from my old blog jaganadhg.freeflux.net