Question Generation using Natural Language Processing - Generate distractors for MCQs

[@PracticalAIbyRamsri]

00:00 Course Introduction 02:00 Course Outline 07:22 Theory - Generate distractors with Wordnet 15:16 Code - Generate distractors with Wordnet 30:14 Theory - Generate distractors with ConceptNet 36:30 Code - Generate distractors with ConceptNet 45:00 Theory - Generate distractors with Sense2Vec 52:32 Code - Generate distractors with Sense2Vec 58:09 Theory - Filter distractors of Sense2Vec 01:00:33 Code - Filter distractors of Sense2Vec

[@arundhathivadla7660]

My best wishes All the best Goutham gaaru

[@RahulCap4]

Great work Ramsri! best wishes for the course

[@PracticalAIbyRamsri]

Thanks Rahul :)

[@anujsingh2455]

Thanks Ramsri for such a great effort and providing such a quality content.. you are gem 🙏🙏

[@PracticalAIbyRamsri]

Thank you Anuj :)

[@niharikareddy4354]

Great Initiative! It is going to help many..keep posting👍

[@PracticalAIbyRamsri]

Thanks Niharika :)

[@totoro_r1668]

Thank you Ramsri! Is it possible to adapt the model for French Question/answer generation tasks?

[@ahsanrossi4328]

one of the best course ever Ramsri. i purchased the course and it is mind blowing. one request to you, is there any course or video how to generate questions from a given context and fine tune the model based on the any dataset ?

[@PracticalAIbyRamsri]

Thanks Ahsan! So this course covers that part right? Generation of questions from text

[@ahsanrossi4328]

@@PracticalAIbyRamsri sorry I may be misunderstood but you can say it clearly, I thought we have answer generation from questions, for example, MCQ, etc.. but seeking help how to generate the question first from the context.

[@ahsanrossi4328]

for example, for the given context, how did you generate the question first "Which is the longest river ?" The model generates random question from context or any other way ... Sorry if i misunderstood.

[@PracticalAIbyRamsri]

@@ahsanrossi4328 In question generation using NLP course on Udemy I teach to generate questions using T5 model

[@ahsanrossi4328]

Sorry Ramsri, i got that part in the course udemy. I was looking into your another course in learn NLP, I purchased both courses... Thanks :)

[@kumarprateek1279]

Thanks man great work. I wish for your subscribers to grow really really high.

[@PracticalAIbyRamsri]

Thanks a lot, Kumar :)

[@ravitanwar9537]

subbed!how to use this api for finding synonyms to a given word?

[@anudeepreddykatta]

Hello Ramsri, If we are trying to automate the process of generating distractors, say we had our answer as USA then by using Sense2Vec and Filtering distractors techniques explained in the video we still had America and United States in the list which is not a great choice to suggest to the user, Is there any way to remove them? ( I tried playing with threshold but it didn't work). Also, Wordnet and Conceptnet will not work well if we had to generate distractors for the river Krishna instead of Nile, What do we do in that case. TIA!

[@PracticalAIbyRamsri]

Hi Anudeep ! Get contextual embeddings and compare. EG: Take the sentence "Joe Biden is president of? USA" get contextual embedding vector using sentence BERT etc. Similarly get for America, Canada, etc. Filter by close contextual vectors. You can give it a try and let me know. 2) If trained on WIkipedia data the results will be generic. SO try other vectors like Google News vector etc if a particular one doesn't yield good results. Or train your own word2vec

[@srikanthporandla4923]

can you share the full code of MCQS generations?

[@anmol23narang41]

Hi How can we automatically find the sense of an option for a given question. For example- In BAT, Is there a way to determine whether it's being referred to as noun or verb with respect to the question? Thanks in Advance

[@PracticalAIbyRamsri]

The topic is Word Sense Disambiguation (WSD). Explained at a later point in the course.

[@PHVijayaRaju]

Can we use for other languages for generating MCQs?

[@PracticalAIbyRamsri]

Hi Joyce, Wordnet and word vectors are available in other languages. So you can use them for distractors using the same algorithm.

[@emmanuelonyeji690]

The udemy course

[@pavithra3148]

Sir I would like to buy the course sir but the price of the Udemy course was 2240 can you please help us by providing coupon code sir .it would be really helpfull for us

[@PracticalAIbyRamsri]

Pavithra ! Try this coupon code valid for 3 days : www.udemy.com/course/question-generation-using-natural-language-processing/?couponCode=BB7AA1D6E91A9C3580DC

[@lowcode-damian]

@@PracticalAIbyRamsri I would be also interested to buy the course :)

[@PracticalAIbyRamsri]

@@lowcode-damian See if this works: www.udemy.com/course/question-generation-using-natural-language-processing/?couponCode=EC7675D158B637BBF0BB

[@lowcode-damian]

Thank you :)

[@MonikaSingh-nu5sg]

Please Share coupon code to me as well .....i really want to buy this course .