Word2Vec : Natural Language Processing

[@matthewson8917]

Finally, someone made a clear and concise introduction for Word2Vec! I admire you!

[@revathik9225]

Great video. One thing to add: instead of always discarding the context vectors at the end, another strategy (mentioned in other videos/articles) is to concatenate or add the two vectors for a word.

[@hahahaYL-h3x]

You are a genius! You are able to explain abstract things well with only white board (no animation!)

[@sefinehtesfa5189]

I appreciate your explanations. I was stuck on Word2Vec. However, you explainde this more than enough. Thank you so much!!!

[@DennisRice-lh3nd]

The best and most concise explanation of Word2Vec that I've seen so far. I probably need to go back and review gradient descent again, because updating the weights is still confusing.

[@DC-tq1kh]

Probably the best video explanation I watched. thank you.

[@DoomScrollSchool]

This is amazing!! I remember trying to understand this back when it was first published, and i failed so hard...thank you

[@jett_royce]

Great video. You mentioned that we discard the context vectors and take the main vectors as our final word embeddings. I just want to add that based on the literature, you can also add the main and context vectors, or concatenate them.

[@drsandeepvm5622]

Exceptionally simplified explanation. Thanks 😊

[@yangwang9688]

Great explanation, it’s pure gold! Can you make a video in terms of hierarchical softmax? It confuses me for a long time.

[@bharathtvadhoola715]

Awesome video! numerical example was particularly helpful. Cheers :)

[@nobodyelse2230]

You're so good! Thank you very much for this explanation, my minds are so slear after it, magic!

[@saikatroy3818]

Explained in lucid manner. Nice video.

[@vctorroferz]

Incredible and amazing explanation! Thanks so much for such great content!

[@ritvikmath]

Glad you liked it!

[@SonalGore-q1v]

Cool....Made it very easy for me to understand about word2vec. Great explanation !!

[@ritvikmath]

Glad it helped!

[@junekang2364]

Great video! Thank you for explaining this so clear.

[@Recoils14]

This explanation is so well done! Thank you!

[@ritvikmath]

Glad it was helpful!

[@MikeKittelberger]

Awesome explanation, you safe me a lot of time. Thank you! :)

[@purpleearth]

This is a superb explanation

[@dominicprior9844]

Wow, such a nice clear video! Many thanks!

[@utuber1752]

EXCELLENT explanation!

[@harshalingutla7318]

Brilliant explanation. To the point.

[@FullStackAmigo]

I learned something, thanks a lot!

[@austossen]

this is by far the best explanation of word2vec on youtube over any university lecture but one question. where and how are you getting your initial vector values in this example for V(like) and V(data)? can you also clarify the components of the vector? you have two columns/elements for each vector in the main vector and context vector.

[@tellyrotsyourbrain]

Exactly what I was wondering. Where do the initialisations of these vectors come from? I'm guessing they're initialised randomly, with the algorithm evolving the distribution into something that relates to the training data.

[@michaelm358]

Outsanding!

[@hameddadgour]

Great presentation!

[@ebujak1]

Great Word2Vec video. Nice touch with blue for Main and red for Context. Working an example with numbers is also very good. Can we please have a follow up video (from ritvik "from the future" :) ) to do the UPDATE part of the loop? We know there are multiple ways to do this UPDATE, but just expound on the "simple" method of moving the vectors as you illustrated in 2D. Thanks.

[@testimonies2914]

Nice ...thannk you so much

[@zhiyili6707]

Thank you for the video! Great content!

[@pampagrimaldi8584]

great explanation! thank you!

[@ritvikmath]

Glad it was helpful!

[@gnkk6002]

Good job as always👏

[@kostjamarschke4613]

Phenomenal stuff!

[@kostjamarschke4613]

Honestly phenomenal. You covered 80% of a 90-minute Masters-level ML lecture in 13 minutes and made it very easy to understand.

[@muhammadal-qurishi7110]

You are always great in your explanations Can you explain CRF please

[@vincentouwendijk3746]

Great video!

[@ritvikmath]

Glad you enjoyed it

[@arshadkazi4559]

Could you elaborate more on context and main embedding please? 6:00

[@limkangwei6339]

This is fascinating!!! Question: How do we get the initial vectors before even starting the for loop?

[@edphi]

Very awesome

[@Halo-uz9nd]

Another fantastic video! The main/context embeddings kind of confused me but the ending really cleared it up. Curious to know if there is a deliberate way of choosing # of dimensions or if its simply trial and error. On a side note, will you be participating in the "66 days of data" challenge this July?

[@gabrielemoro304]

the dot product between V_like and V_data should be 0.4 and not 0.6 right? (min 8:27)

[@KD4wg]

yes, the dot product is 0.4. But our score is the sigmoid of the dot product. And sigmoid(0.4) is roughly 0.598 ^^

[@HiteshKumar-bk2od]

Can you please make a video on transformers, attention and Bert models in detail . It will be in Continuation with word vectors.

[@VincentVanZgegh]

Great video ! Thx

[@ritvikmath]

Glad you liked it!

[@houyao2147]

very intuitive!

[@sa-vx5vi]

Thanks a lot

[@mmenjic]

And how do we distinguish between 2 same words in this case "data" if we have larger text where data is next to science but it is also close to other word or words for example data processing, data mining, data validation, data.... data..... everything, is it the case that word data is "closer" to word science in general than it is to word processing or there is some other mechanism for this ?

[@oraz.]

What determines the dimension of the word vectors? Edit I guess it's just an engineering decision.

[@0815-j3s]

Simply since the distances are invariant under orthogonal transformations of the space "embeddings" (by the way: Not in the mathematical sense) are not unique. But they also depend on the prupose of the model.

[@marsbux-tech]

Rather than using iterative approximations, can we not use a function to determine the final vector for each word in just one step?

[@priyasah6123]

How to calculate score ? We need to take dot product of what exactly with main and context ?

[@kundansaha4877]

please make a video how spacy works

[@bartoszko4028]

Well explained, but doesn't sigmoid function give values from 0 to 1? Then maybe it should be tanh activation function?

[@ebujak1]

Score needs to be positive since the error needs to be opposite (error = label - score). tanh is negative and positive ... so hard to interpret error.

[@mosca-tse-tse]

So how does it happen than king - man + woman = queen, a very common selling example of wod2vec?

[@muhammadal-qurishi7110]

check this video and you will have clear understanding kzbin.info/www/bejne/al60oICaiNZ6las

[@mosca-tse-tse]

@@muhammadal-qurishi7110 cool 👍🏻 thanks!

[@lsacy8347]

why is this word2vec so different from "conventional" word2vec where you use a neural network and bag of words to calculate the weights matrix? ok, my bad, just realized you talked about neural network update method towards the end.

[@jagadeeshwaran1512]

How does the Main embedding is calculated?. How is the vector defined?

[@moatasem444]

There is a prepared embedding?

[@golnoushghiasi7698]

Can you explain some more NLP models such as Bert, Fasttext and transformers?

[@nobodyelse2230]

Hello again! After a couple of days of studying another materials, I realized that I don't understand where the formulas for the score and the error come from, I don't see them in the books I read (and I haven't seen this approach with labels yet either). Are there any papers or books where I can find them and the prove that they work?

[@bilalsidiqi9992]

Could you please implement some NLP models in pytorch or tensorflow?

[@agamchug595]

How is the initial size of the vectors determined? As you mentioned that traditionally there may be vectors with 50 dimensions and your example has used 2 to ease understanding.

[@CarlJohnson-jj9ic]

Is chloroform FDA approved for a mechanics lean on a judgement?

[@sachinrathi7814]

Can someone will explain where those 2 numeric values (vector representation) for like word, -.2 and -1 comes from ? Word2vec is used to create numeric representation of words in vector form or create vector in such a way that word which are close in documents are also close in vectors representation as well ?

[@roysarit]

initially the values are randomly set. after each iteration of computation, the values keep getting updated.

[@vyvu443]

Ritvik... can you do NLP, word embeddings with GloVe and FastText :D :D :D Thank you in advance!

[@kevinscaria]

Wow

[@RonLWilson]

Just FYI, I just made and uploaded a KZbin video where I used bits of your video here as a use case to illustrate a graphical ontology language that I am calling UniML, Universal Modeling L (not to be confused with UML, Unified modeling Language). While things like neural nets map transformation or functions (f: X -> Y) UniML attempts to model more than that and model the thing itself, including any such transformations as an ontology model but not using just symbols such as does ontology languages such as OWL or RDF but graphically so that graphically depicts the under lying data structures. Here is a link to my video kzbin.info/www/bejne/mJrSq2Voa6iWjbs

[@gagangayari5981]

Perfect explanation. How would doc2vec fit along the same lines with the algorithm you mentioned ? Can you please briefly tell ? @ritvikmath

[@shakeelm5339]

I have no clue what the heck he talked about.