N Grams Models Applications


too many applications of n gram models
in speech processing and in text processing
let’s visit some of them the first very important application is in OCR OCR
means optical character recognition like you use an image of text and then you
convert that image into a real text like you scan a news paper and then from
this scanned copy of newspaper you produce actual text
that’s done by the optical character recognition so while scanning a
newspaper or scanning a book if you find some word is missing or confusing
that missing word is very easily predicted by the n-gram models
similarly if you use the handwritten text for scanning the
handwritten text is normally confusing so for that scenario it’s also very
helpful. The second application is correcting a sentence when we write a
sentence if there is some spelling mistake it is very easily identified by
all the word processors like ms word and others word processing applications but
if the Spelling’s are correct but the word is not placed rightly then it is
there is a some problem so for example like if we write deer sir instead of
dear sir this one so here this deer and this dear both are right words in the
dictionary so there is no spelling mistake here so the normal word
processing application will not identify any error but if you use an n gram
model the n gram model will will suggest to us that sir is normally followed by
this dear not this deer so it will notify us that we
should correct this this phrase this is an incorrect phrase so it’s very helpful
in such scenario and the next is speech recognition there
are many applications these days I think all the Android and mobile smartphones
have this feature that you speak and your speech is converted into
text and many apps which can do this even Google offers speech
recognition feature in internet search so you can use this elegant model in speech
recognition as well so because there are many words in English and I think in almost a all
the languages that have the same sound but actually they are very different for
instance we have the word I and eye so these have say almost same sound but but
they have quite different meanings so with the same sound you have to predict
what is the right word in the sentence this is done by n gram model like
if someone says I am eating the speech recognition application will not write
Even eye here but it will write this this one here because it is it comes in the
start of the sentence and normally am is followed by this I not this eye so from
the n gram model we can easily identify the right word in the given police of
the sentence where two words have same sounds similarly if someone says his eye
got hurt here we will use this eye that the part of the body so the next
application is in translation in translation
there are, we usually find very funny scenarios when we try to translate a
sentence word by word for example if we translate a sentence from national
language of Pakistan like he is the biggest minister then it will become he’s
the biggest Minister of Pakistan and so it is very funny translation but whereas
if we consult with n gram model the The n gram will suggest us that
we should we should use here the Prime Minister of Pakistan
so apparently Prime and the biggest are not much different they they are almost
synonyms minister
what is mostly followed by the prime not the biggest word so we will use instead
prime instead of biggest so these are few applications of these n gram
models. there is also an other application that is the suggestions while
typing so this scenario is you have observed I think all of you have
observed it while typing on your mobile phones in mobile phone when we try to
type some word . similar words are suggested these words are suggested
depending upon two things the first thing is the spelling of the word and
the second thing is depending upon the n-gram model because our smartphone
is automatically trained according to our texts text messages so on the basis of the ngram model it suggests the right word to be typed so these are some of the applications of n gram models and there
are many other applications as well. or the whole you can say the anagram
models are very important and you can say the building blocks of the speech
processing

1 thought on “N Grams Models Applications

  1. Hello Sir,
    Very good explanation of OCR. I have a question regarding how can one get started with programming n grams in Java?

Leave a Reply

Your email address will not be published. Required fields are marked *