Next Generation Writing With AI and ML

Introduction

In today world writing skill become more and more important every day to day life it’s important to have a good presentation skill so that your

content shine different from another person but this thing is someone wo can master in one or two days it takes time, effort, and experience. content writing is vast field in it self and every big or small company need content writer to write blog and article for their website because good content is easily understood by everyone.

But there is a big gap between a technology person and content writer someone with good technology knowledge is not necessary able to express their thoughts to other and this can lead to very wide communication gap due to that communication gap he/she may not be able to express their idea and unable to achieve that breakthrough he/she may deserved this not an hypothetical situation this thing happen to many person nowadays and its still going on the other hand person with good writing skill not necessary enough technical knowledge to write a article/blog this lead to miscommunication and this may ultimately lead create a myth in the market and this myth lead to thing that may lead to may technology not become successful that may intended to and thing I talk till now they are not hypothetical situation they are happening to many person to real life so what is the solution for this one of the most common solution that going from may year is to hold meeting between technical writer and technology person but this process is very time consuming and not always that much helpful is many cases when both writer and technology guy is not experiences this process become very typical and very time consuming and this always lead to miscommunication

To solve this kind of problem a startup named Grammarly use machine learning and Artificial intelligence to provide a enhance writing assistant that helps you to writing

Feature of Grammarly

Grammarly support many feature some of them are listed below:-

• AI based assistant that helps you writing and find spelling mistake as you type
• Helps you to find grammatical mistake and suggest to the write way to write that sentence, paragraph and line.
• Helps you to find attractiveness of your passage
• Helps you to choose write word that fit in the passage and key word also
• Give you the tone of the article (aggressive, formal, informal etc.

How grammarly works

Exactly how grammarly work its a very long topic or you may say that its a degree course in itself but in this upcoming article I am going to tell you the very high level how it works

• Autocorrection
• Pattern recognition
• Natural Language processing
• ML/AI

Auto correction

Grammarly first and foremost use auto correction for this grammarly has the huge database that content nearly all the words for english and words that people generally used and they are not belong to english language for eg:- company/organization/brands names, fictional character, commonly used short form, name of famous person or celebrity etc there is never ending list of words in grammarly database

Whenever someone start typing let say he/she type H grammarly load all word that start with H and then he/she type let say EL so now grammarly has all the words that start with HEL so what is next alphabet if I ask a human he/she might say i am going write help, hello for a single human answer is directly proportional to the number word he/she know that start with hel not more than this if he/she didn’t that words exits this is reason why grammarly collect that many words in their database and their database is updated regularly

Pattern recognition

So what is pattern recognition and how grammarly used this concept to understand this term much better let me ask you a simple fill in the blank

1. Somebody ___________ me
2. _________ how are you

There is 90% of chance for the first blank you fill Help and for second blank you fill Hello not you more than 98% of people fill the same thing.

So why this happen; from the birth till the day we die we encounter many sentence daily whether speak by someone from friends and family or we form someone we don’t know or may be we read somewhere all that time whether we realize or not our brain start making pattern.

Grammarly work on the same principle grammarly also created that pattern in their mind (in computer world correct word is database) and use them to correct things. So how grammarly created pattern and how he use that pattern to guess things this is the real challenge and this is all MI/AL about.

Natural Language processing

Natural Language processing(NPL) many person said NPL is just a brach ML/AI in which computer that content reading the text or interpret what human is talking but this is wrong this just a small branch of NPL which in technical term they are know as Speech Synthesis and Speech recognition

So what in NPL its umbrella term that consists of many sub but interrelated topic Speech Synthesis and Speech recognition are just one of them some of the term that also comes under NPL are given below:-

• OCR(Optical character recognition) which defined by simply looking at the word or character computer tell what what word it is, whether it comes from image, hard paper, pdf or doc, or appear in video anywhere if A is written computer said it’s a A.
• Understanding standing the tone of the sentence not every sentence is same how computer understand the tone of the sentence whether the sentence is formal, informal, aggressive, requesting sentence etc if computer understand right tone of the sentence computer was able to give much better answer to he/she.
• Feeling/Emotion from the line or paragraph, whenever human read a book/novel he/she was easily able to understand the emotion of the character just like they are experience it personally many writer take this and able to deliver the very good novel/book/comic anything you say how computer was able interpret the same emotion that human was able to understand.

They are not all the field that comes under NPL they are the field that i find most interesting

I am not saying grammarly using every bit of NPL but NPL is the core concept that grammarly use

ML/AI

ML/AI (Machining learning and Artificial Intelligence) both of them was separate branch in themself but it's always good to use ML/AI with each other but why

To understand this let take an eg of sprinter and marathon runner.

“So what is the difference both of them the way they run sprinter was trained to use their all the power for very short distance for eg 100m while marathon runner are trained to use their power for very long time for eg 10km they have different style of running they master it by hard training and use it in real competition. ”

The key word here are training and competition and in computer language they are known as ML/AI

So if we train like very very hard and never use this anywhere they are just the waste of time and money and some we directly and run in competition and never train in our whole life we can say this their no chance that we are going to win competition.

Same with computer training means ML and competition means AI that why is always good say them together.

Same way grammarly work they use ML/AI to their data and use them i can’t say directly what way they(in ML/AI computer world the correct word is model) they use because there is N number of model some of them are public available and some they made by themself they are private to them just like in our sprinter and marathon runner example some training are common for both of them, some of them is unique to their field and some them is specific to training center where they are training.

Concussion

In conclusion grammarly is one of the many example that understand and use the concept of ML/AI and created the very powerful tool that helps many people in the world to do their day to day activity

Thank for everyone to read my article till end if you have any doubt please comment if you have any suggestion please mail all comment both positive and negative is more than welcomed.

Contact Detail

LinkeDin [https://www.linkedin.com/in/suyash-garg-50245b1b7]

Additional Tags

#arth #vimaldaga #machinelearning #righteducation #rightmentor #worldrecordholder