Yesterday, there was ‘social’ networking reception. and Mike Trick tweeted about the reception as following.
“Big crowd at suicidal networking reception at #informs2015 . Should have checked twitter handles at door. ” @miketrick
We all know what happened. Autocorrection on smartphones.
I also have enough incidents to hate autocorrect on my iPhone.
How the autocorrect works?
I found it relies on the simple but famous Bayes’ theorem.
argmaxc P(w|c) P(c)
P(c), the probability that a proposed correction c stands on its own.
P(w|c), the probability that w would be typed in a text when the author meant c.
This problem can be boiled down to how to get the right probabilities or how to construct personalized probabilities.
It is also related to Analytics, big data, and obviously the natural language processing.
It might be an interesting topic to take a look.
However, do we really have to improve autocorrect on our smartphone so that it takes away unexpected chances to laugh together? I doubt.
What do you think? Especially who were in ‘Suicidal’ networking reception.
P.S. Mike, I am sorry that I make fun of your autocorrect, the second time.
S. Phil Kim (@ksphil)