Stamenov/twittersentimentbugs

Uppercase/Lowercase handled in a confusing way

Opened this issue · 0 comments

What steps will reproduce the problem?

Search for the same term once Capitalized and once lowercased. And add them
to your list of saved searches.

E.g. 
Searching for "Kites" (a recent Hindi movie). Add it to your list of saved
searches.
Now, search for "kites" and add it to your list of saved searches.

What is the expected output? What do you see instead?

They either should be treated as separate entities or the same
(case-insensitive). But the current system seems to be doing a mixture of
both. Searching for "kites" has less than 50 tweet results - though the
tweet themselves include all possible versions of the word - "KITES, kites,
Kites". Searching for "Kites" has 1000s of tweets as results, again
including the word in all possible cases.

Moreover, even when you search for "Kites", once the webpage has refreshed
after fetching the results, the search bar text automatically becomes
"kites" and that's what shows up above the graph too (i.e. Sentiment
results about "kites").

Also, once you have stored "kites" as a saved search, you can no longer
save the capitalized version "Kites" in your list. 

Original issue reported on code.google.com by ktic...@gmail.com on 22 May 2010 at 4:53