hi, your video is good and very useful for me. can you please provide how to get file from twitter and iam new to work on R software. if you don't mind i need some basic videos step by step process from getting data and doing wordcloud and sentiment analysis on R. i am waiting for your reply for example i need a data from twitter big data in agriculture. like that i want to get data sheet from twitter and run the word cloud and sentiment analysis on R. can you suggest me
Thanks so much for this detailed explanation. Please I'm having an issue fixing the TermDocumentMatrix error. I get this error when I get to the tdm: Error in nchar(Terms(x), type = "chars") : invalid multibyte string, element 220 How do I fix it?
Hello Sir, I have imported all the mentioned libraries, but when I run the command: get_nrc_sentiment, I am getting the error as: Error: 'vec_as_subscript' is not an exported object from 'namespace:vctrs' Run `rlang::last_error()` to see where the error occurred. Can you please help
Sir, unable to perform sentiment analysis due to unavailability of package syuzhet and lubridate.... others like ggplot,scales are there... R is showing error displaying there is no packages named syuzhet and lubridate... what to do Sir.. plz help..
I am getting the below waring messages, how can I ignore this Warning message: In tm_map.SimpleCorpus(corpus, tolower) : transformation drops documents
Many thanks for your video. Can you please tell us something more about the Datasets Apple & Apple 2. Did you export it from Twitter Analytics/Activity.
It may happen if data size is to big to be handled by your computer. Try reducing data size. Or you can try google colab: ua-cam.com/video/XVfn6IpoUPU/v-deo.html
Hey, i super appreciate to your work it easy to understand, but i have questions is that code can be implemented on any language such as Bahasa Indonesia? Thanks you, if you are available for reply
hi Sir the way u removed stocks with stock i faced this challenge with multiple words like google was like goog googl, etc, and few other words like microsft was written twive as microsoft and msft, amazon as amzn. How to automate this process of replacing these kind of words
Thanks Bharatendra sir for sharing your knowledge.IT has been very helpful in my current project. We have below list of sentiments packages. #syuzhet #SentimentR #Rsentiment #Sentimentanalysis I am still validating all of above. Please let us know your thought about package recommendation for sentiment analysis.
Thanks sir for your confirmation. yes Syuzhet is giving expected result for emotion and sentiments with some misclassification. I am using SentimentR to get avg sentiment by respective dimension and to get positive word list and negative word list.
It was a Great Lecture and very easy to understand. Thanks for this video. Can you please tell me how did you pull tweets after earnings. which part of the video or lines we need to follow to pull the tweets? I have followed this complete video and got the results. I want to extract Preprocessed csv file..please help me by your suggestions? Is this preprocessed file stored in R library of my computer?
What version of R have u used for the entire project as I have having difficulties with R 3.5 and R 4.0 for installing some packages like Corpus and iconv
Let's say your twitter data is in apple. Let's say you have all sentiment scores in SCORES. Then you can combine them using cbind(apple, SCORES). It will combine the columns.
what should users of window use in the place of utf-8-mac at 1:50? I have tried to reproduce your example but it is not working, sorry I am newbie to these things, thank you in advance!
Dear Bharatendra Rai sir you are doing great job with these videos, very informational. Please guide how to get data from twitter or other social media handles for a specific word. For example real estate market's particular project in Chennai. Thanks for your valuable inputs.
@@bkrai thanks i will try and get back to you my question is if i want to know what are kolkata people tweeting on one specific tweet how may i drill down to location and then to that #subject
Fantastic video! Surprised I have not run into your channel before. I am analyzing customer survey results from different countries, and one problem that I am having is that some of the countries have very polite people (not a bad thing of course) and their survey results show positive, when in fact they are not happy with the result. Have you had this problem before, and if so how did you handle it?
Sir, i ran Library (tm) also on window but getting Error in iconv(apple$text, to = "utf-win") : unsupported conversion from '' to 'utf-win' in codepage 1252. pl help me .
Awesome video. I see the video was uploaded around 1 year back. Is there any new library after that which can further simplify the coding ? Also will there be further simplification of the code if we use Microsoft Cognitive Services (Text Analytics API) .
If you are doing sentiment analysis, the code has just one or two lines. When using other tools such as MS cognitive services, note that the results are likely to be different.
thank you for video. but lettercloud function is not working. R is not send message "lettercloud is not working" but there is no graph. why please help me?
@@bkrai hi again. i fixed the problem finally. firstly run the code "lettercloud" and there will be an empty page at the corner and select "open in the browser" upside. and again there will be an empty page and finally click f5 for the refresh the internet page. so you will see the lettercloud. :)
Congrats on the work @@bkrai. Regarding lettercloud, after I run the code, it only responds with one word, but it works fine for wordcloud2 to make stars or circles. Do you know what the problem is?
This video is really helpful, thank you very much! but how can we evaluate the performance of the sentiment analysis you have just made? Are there any metric to use? Please help...
One way could be to score tweets manually and then compare the results from automated sentiment analysis. And then mismatch % could work as a metric to assess the performance.
Could you please help me out... I was unable to get the letter cloud. All steps followed as mentioned in the video. The video was very helpful in all the other aspects.
Hello sir.. thoroughly enjoyed the video, and was trying to replicate the same in windows, but was getting some errors as in fourth line there were some accented characters which i could not get rid of - "RT @SylvaCap: Things might get ugly for $aapl with the iphone delay. With $aapl down that means almost all of the FANG stocks were down pos… " the letters after pos in the end.. (i am thinking this might be due to some encoding issue, I am using utf-8 as suggested by you in one of the comments below), i tried using your code as well, still the result were same.. Can you help me on how to move forward? Thanks
I am getting this error Error in nchar(Terms(x), type = "chars") : invalid multibyte string, element 22, when I write this code for tdm. Can you please help me with code error?
Hi! Thanks a lot for your video! I have a question. How do I remove stopwords for German? Is it the same line of code with putting german instead of english? Thanks
Hey bharat, Thank you very much for such a nice explanation. Need your help. am not able to see the output of letter cloud as shown. Please suggest. Regards, Murali
Thank you for your reply. Actually, am using the same code for lettercloud but the output is not same as yours. code: "letterCloud(w,word = "apple",wordSize = 1)" The output is apple with black letters. Please share your email id so that I can share the output
Hi! I have a question While trying to plot the wordcloud I get this message: hertha could not be fit on page. It will not be plotted. 6: In wordcloud(words = names(w), freq = w, max.words = 100, min.freq = 5, : Any way I can fix it?
For the Sentiment bar chart comparison, It would be an easier comparison if the bars were based on % of total row count instead of comparing counts. Would you do rowsum(x) / sum(x)? Or is it colsum(x) / sum(x)
For converting y-axis to % you can use: barplot(100*colSums(s)/sum(s), las = 2, col = rainbow(10), ylab = 'Percentage', main = '% Sentiment Scores for Apple Tweets')
Mr.Baharat, After getting text (tweets) and clean them, How can we rewrite them as a .csv file to get the sentiment analysis of the new data with the clean text. What you did here is that you analysed the raw data with unclean text? Thanks
hello sir, i have a doubt, for the first file apple.csv you have done all the pre processing(cleaning of data) ,but for the second time when you are performing the actual sentimental analysis you have not done any pre processing. So you are doing the sentimental analysis without any cleaning in the second file?
Pre processing helps when you are developing wordcloud as you don't want to see thing that are not relevant. However, for sentiment analysis it may not matter if common words like 'the' or 'is' are there or not.
if text form is in different language other than english how can i solve such of text. any video related to NLP bigram,wupalmer, stemming, lemmatization
Really excellent video!! Where can we find dictionaries with keywords annotated with sentiments (e.g. ugly as you mentioned in video)? Also, how can we add sentiment dictionary for new language? Any video on these topics? Thank you very much ... :)
Mr.Bharat, when i want now to pull new data about apple from Twitter , R returns the tweets dated 2018. My question is: how can i pull tweets ,for example, about the second quarter 2016? Is there any way to specify an exact date range to pull tweets? Regards.
Hello sir, I am having this issue while using the code for twitter and no idea on how to resolve it. > setup_twitter_oauth(api_key,api_secret,access_token,access_secret) [1] "Using direct authentication" Error in check_twitter_oauth() : OAuth authentication error: This most likely means that you have incorrectly called setup_twitter_oauth()' > setup_twitter_oauth(api_key,api_secret,access_token=NULL,access_secret=NULL) [1] "Using browser based authentication" Error in init_oauth1.0(self$endpoint, self$app, permission = self$params$permission, : Forbidden (HTTP 403). You guidance will be of great help.
hello sir thanks for such a knowledgeable video sir I want to know that which year of data you used in this video and may we use this data for research purposes.please reply thanks again.
Hello, I have a question, I would like to ask how could I view an excel file with Chinese characters. After I str my data, it only shows sth like this: U+84EE. I have tried this corpus
Hi Bharat; When looking at the data file, we can see that most of columns are N/A. How we can get Longitude and Latitude for example? what is the meaning of replyToSN and replyToSID? you have mentioned that Apple two datasets are taken from Twitter before and after quarterly earnings report. Could you please let us know what quarter is that? I mean is it after the 1, 2, 3 or fourth quarter the data was taken???
Longitude and Latitude is only available for those Twitter users who agree to share their location. Due to privacy reasons many people do not like to share their location and that's why majority of tweets do not have this info. replyToSN - reply to Screenname replyToSID - reply to sender ID apple2.csv has data for 2nd quarter of 2017.
Hello sir, Don't we have to pre process the text before running get_nrc_sentiment, you are directly passing the comments into that command, but in case of wordcloud you are removing stopwords, urls etc...Could you pls explain why are you doing that?
For sentiment analysis words with no emotions will automatically be ignored. In wordcloud we do not want those words to appear, that are not meaningful.
Too good Video....your explanation is awesome sir....Thank you so much
You are welcome!
Can you please help, after doing text preprocessing I am trying to perform tdm however getting bellow error.
dtm
By far one of the best teacher of Data Science, who believes in keeping it simple. Absolutely love them.
Thanks for comments!
hi, your video is good and very useful for me. can you please provide how to get file from twitter and iam new to work on R software. if you don't mind i need some basic videos step by step process from getting data and doing wordcloud and sentiment analysis on R. i am waiting for your reply
for example i need a data from twitter big data in agriculture. like that i want to get data sheet from twitter and run the word cloud and sentiment analysis on R. can you suggest me
You can try these playlists:
ua-cam.com/play/PL34t5iLfZddt0tt5GdDy3ny6X5RQvwrp6.html
ua-cam.com/play/PL34t5iLfZddtfgOcE6aKCEXWtCkolpgjV.html
Sir, I have a problem. wordcloud2 not working at my pc.
error message:
Error in gsub(
For pc make sure you use "utf-8" in line-7 and not "utf-8-mac".
@@bkrai I am sure. I use to gsub func for turkish character to convert.
Example:
corpus = gsub("ÅŸ", "s", corpus)
ş > s
@@bars2251 Error in gsub("
Thankyou for the video.....How do you create apple.csv file in your desktop that contains 1000 tweets?
I got it from Twitter:
ua-cam.com/video/QETCjkQ3CBw/v-deo.html
@@bkrai Thank you sir
You are welcome!
Thanks so much for this detailed explanation. Please I'm having an issue fixing the TermDocumentMatrix error. I get this error when I get to the tdm:
Error in nchar(Terms(x), type = "chars") :
invalid multibyte string, element 220
How do I fix it?
Difficult to say much without looking at the code.
>>>corpus
If you are using Windows computer, use: "utf-8" instead of "utf-8-mac".
Hello Sir, I have imported all the mentioned libraries, but when I run the command: get_nrc_sentiment, I am getting the error as:
Error: 'vec_as_subscript' is not an exported object from 'namespace:vctrs'
Run `rlang::last_error()` to see where the error occurred.
Can you please help
Check structure of your data and make sure it is in right format.
Sir, unable to perform sentiment analysis due to unavailability of package syuzhet and lubridate.... others like ggplot,scales are there... R is showing error displaying there is no packages named syuzhet and lubridate... what to do Sir.. plz help..
You need to install those packages first.
I am getting the below waring messages, how can I ignore this
Warning message:
In tm_map.SimpleCorpus(corpus, tolower) : transformation drops documents
You can check the matrix and see if you are really losing anything.
Many thanks for your video. Can you please tell us something more about the Datasets Apple & Apple 2. Did you export it from Twitter Analytics/Activity.
I used this:
ua-cam.com/video/QETCjkQ3CBw/v-deo.html
Warning message:
In tm_map.SimpleCorpus(corpus, tolower) : transformation drops documents..
Message is shown
Warning message in R are ok. You should not get error message.
Sir while trying this example my TermDocumentMatrix is not working...system gets hanged when the function is called
It may happen if data size is to big to be handled by your computer. Try reducing data size. Or you can try google colab:
ua-cam.com/video/XVfn6IpoUPU/v-deo.html
Sir how can cav file be generated?? Although you provided Cvs file but how can I extract file from twitter or facebook
For getting Twitter data, see if this helps:
ua-cam.com/video/bxJseJ-aZJk/v-deo.html
Hey, i super appreciate to your work it easy to understand, but i have questions is that code can be implemented on any language such as Bahasa Indonesia?
Thanks you, if you are available for reply
I'll look into it.
hi Sir the way u removed stocks with stock i faced this challenge with multiple words like google was like goog googl, etc, and few other words like microsft was written twive as microsoft and msft, amazon as amzn. How to automate this process of replacing these kind of words
You can do it for just top 2 or 3 words. After that it doesn't really matter.
Thanks Bharatendra sir for sharing your knowledge.IT has been very helpful in my current project. We have below list of sentiments packages.
#syuzhet
#SentimentR
#Rsentiment
#Sentimentanalysis
I am still validating all of above. Please let us know your thought about package recommendation for sentiment analysis.
syuzhet should work fine for sentiment analysis.
Thanks sir for your confirmation. yes Syuzhet is giving expected result for emotion and sentiments with some misclassification. I am using SentimentR to get avg sentiment by respective dimension and to get positive word list and negative word list.
It was a Great Lecture and very easy to understand. Thanks for this video.
Can you please tell me how did you pull tweets after earnings.
which part of the video or lines we need to follow to pull the tweets?
I have followed this complete video and got the results. I want to extract Preprocessed csv file..please help me by your suggestions?
Is this preprocessed file stored in R library of my computer?
For pulling tweets you can follow 1st video in this playlist:
ua-cam.com/play/PL34t5iLfZddt0tt5GdDy3ny6X5RQvwrp6.html
The best explanation and walk-through of sentiment analysis I've seen so far. Well done, very helpful.
Thanks for the feedback!
What version of R have u used for the entire project as I have having difficulties with R 3.5 and R 4.0 for installing some packages like Corpus and iconv
Probably it has more to do with the computer. I used Mac and that's why used "utf-8-mac". If you are using windows, make sure to use "utf-8".
what do i do if i want to save the score in that csv file? or replace the reviews with the score in the file? Please Help.
Let's say your twitter data is in apple. Let's say you have all sentiment scores in SCORES. Then you can combine them using cbind(apple, SCORES). It will combine the columns.
Dr. Bharatendra Rai will try this. Thank you so much
welcome!
Sir when I am going to remove stopwords , still stopwords are present. Plz guide me how to remove these stopwords
It should work fine. Review your code again.
what should users of window use in the place of utf-8-mac at 1:50? I have tried to reproduce your example but it is not working, sorry I am newbie to these things, thank you in advance!
for windows use utf-8
@@bkrai thanks a lot for prompt feedback!
Thaks Bharatendra, your videos are great full. Do you know how a I coud be specify another idiom (like Spanish) for the sentiment analysis?
Dear Bharatendra Rai sir you are doing great job with these videos, very informational. Please guide how to get data from twitter or other social media handles for a specific word. For example real estate market's particular project in Chennai.
Thanks for your valuable inputs.
you can use a word or phrase, just as we use when we do google search.
@@bkrai thanks i will try and get back to you
my question is if i want to know what are kolkata people tweeting on one specific tweet how may i drill down to location and then to that #subject
@@bkrai thanks
you are simply great. i tried and it worked really well. Thank you so much for clear explanation
Thanks for your comments!
Hello sir,
What does 'utc-8-mac' mean and what does it do? Could you please explain?
Thank You!
See it now, probably you already figured it out.
Whichever topic I take to study in R software, I will find your video teaching most interestingly and easily, thank you so much, sir
You are most welcome!
Fantastic video! Surprised I have not run into your channel before. I am analyzing customer survey results from different countries, and one problem that I am having is that some of the countries have very polite people (not a bad thing of course) and their survey results show positive, when in fact they are not happy with the result. Have you had this problem before, and if so how did you handle it?
I've not come across this, but seems an interesting problem.
Sir, What is the Algorithm used here?
It uses nrc lexicon.
Thanks for your knowledge Share. Your explanation helps me a lot to achieve my goal. However, when I ran the scripts
corpus
Check if you ran library(tm) line or not.
Sir, i ran Library (tm) also on window but getting Error in iconv(apple$text, to = "utf-win") :
unsupported conversion from '' to 'utf-win' in codepage 1252. pl help me .
even corpus function is also not working
utf8text
Absolutely awesome video. Leart from your video more than my expectation. Clear and sensational! Thank you.
Thanks for comments!
Awesome video. I see the video was uploaded around 1 year back. Is there any new library after that which can further simplify the coding ? Also will there be further simplification of the code if we use Microsoft Cognitive Services (Text Analytics API) .
If you are doing sentiment analysis, the code has just one or two lines. When using other tools such as MS cognitive services, note that the results are likely to be different.
When I write this code corpus
If you are using windows computer, use "utf-8".
@@bkrai - Oh okay. Understood. Thank you.
Welcome!
Hi Sir in TermdocumentMatrix ---Cleanset is not coming ,do we need to install any separate package for it?
Make sure in TermdocumentMatrix, 'D' is uppercase like 'TermDocumentMatrix.
@@bkrai done sir
Sir please make a video on topic modelling of Twitter data on R🙏
Thanks for the suggestion!
thank you for video. but lettercloud function is not working. R is not send message "lettercloud is not working" but there is no graph. why please help me?
I also noticed it. Looks like this feature doesn't exist anymore.
@@bkrai hi again. i fixed the problem finally. firstly run the code "lettercloud" and there will be an empty page at the corner and select "open in the browser" upside. and again there will be an empty page and finally click f5 for the refresh the internet page. so you will see the lettercloud. :)
Congrats on the work @@bkrai. Regarding lettercloud, after I run the code, it only responds with one word, but it works fine for wordcloud2 to make stars or circles. Do you know what the problem is?
This video is really helpful, thank you very much! but how can we evaluate the performance of the sentiment analysis you have just made? Are there any metric to use? Please help...
One way could be to score tweets manually and then compare the results from automated sentiment analysis. And then mismatch % could work as a metric to assess the performance.
Thank you very much.
This is simply awesome and so much interesting. Loved it. Thank you so much Sir.
Thanks!
Could you please help me out...
I was unable to get the letter cloud.
All steps followed as mentioned in the video.
The video was very helpful in all the other aspects.
I too noticed that lettercloud is no more functional. I hope they update the package soon.
Wow!
I didn't expect such quick response..
Thank you so much for the clarification..🙏🏾
You are welcome!
Thank you soo much for the video but I just wanted to know which algorithm is being used here
For which output?
for apple.csv
Hello sir.. thoroughly enjoyed the video, and was trying to replicate the same in windows, but was getting some errors as in fourth line there were some accented characters which i could not get rid of -
"RT @SylvaCap: Things might get ugly for $aapl with the iphone delay. With $aapl down that means almost all of the FANG stocks were down pos… "
the letters after pos in the end.. (i am thinking this might be due to some encoding issue, I am using utf-8 as suggested by you in one of the comments below),
i tried using your code as well, still the result were same.. Can you help me on how to move forward? Thanks
Sir can you provide the same code for windows users since I'm running the code on windows machine. Because some commands are different in Mac laptop.
Did you try this: ua-cam.com/video/BPjgwdqHM8g/v-deo.html
@@bkrai thank you so much Sir.
You are welcome!
Thank you for your video! It is very helpful and your explanation is so clear! Thanks again.
Thanks for comments!
Thank you Dr. Rai, very informative video kind sir
You are welcome!
I am getting this error Error in nchar(Terms(x), type = "chars") : invalid multibyte string, element 22, when I write this code for tdm. Can you please help me with code error?
This video is very helpful and could explain z test on r
I've used z-test in this:
ua-cam.com/video/oxRy2DMrOF4/v-deo.html
This was really helpful, but i was wondering how to do normalization of words, here you did only for one word, what if i need to do for many words?
For 'normalization', which line of the code in the video are you referring to?
> tdm tdm
Non-/sparse entries: 612739/3617891553
Sparsity : 100%
Error in nchar(Terms(x), type = "chars") :
invalid multibyte string, element 13087
> tdm
Probably 27 GB is too big for your computer.
thank you sir! amazing video. You explanation and pace of videos are on point!. thanks again.
Thanks for comments!
You're a legend! thank you for this video.
You're welcome!
Hi! Thanks a lot for your video! I have a question. How do I remove stopwords for German? Is it the same line of code with putting german instead of english?
Thanks
Hey bharat, Thank you very much for such a nice explanation.
Need your help.
am not able to see the output of letter cloud as shown.
Please suggest.
Regards,
Murali
Difficult to much without looking at codes. Probably size may be too big and doesn't fit in the area available.
Thank you for your reply.
Actually, am using the same code for lettercloud but the output is not same as yours.
code: "letterCloud(w,word = "apple",wordSize = 1)"
The output is apple with black letters.
Please share your email id so that I can share the output
Very good explanation! Excellent job! Btw I have a question about word cloud: does the position of the words mean any relationship?
Position of words is random.
at 7:03 Error in nchar(Terms(x), type = "chars") :
invalid multibyte string, element 19-----Cant find solution anywhere
problem is, you don't have a Mac, so you never converted your Apple text into UTF-8.
Use:
apple$text
@@baruchschwartz819 thank you so much sir. help appreciated
@@marshalmashiat cheers
Thanks!
Excellent Video sir, very elaborate. a quick question.. what would be the windows equivalent of the line tweets
tweets
@@bkrai sir this is showing that unsupported conversion from ' ' to 'utf-8' in codepage 65001 , plz help with this
Names on x-axis and count on y-axis of barplot didn't appear while executing the barplot.It only shows the 📊 itself...please help me out
Excellent one Sir.... Please keep sharing such videos ....Your way of explaining things is Awesome...
Thanks for comments and feedback!
Hi! I have a question
While trying to plot the wordcloud I get this message:
hertha could not be fit on page. It will not be plotted.
6: In wordcloud(words = names(w), freq = w, max.words = 100, min.freq = 5, :
Any way I can fix it?
You can control how big or small a word looks using scale function as shown in the video.
For the Sentiment bar chart comparison, It would be an easier comparison if the bars were based on % of total row count instead of comparing counts.
Would you do rowsum(x) / sum(x)?
Or is it colsum(x) / sum(x)
For converting y-axis to % you can use:
barplot(100*colSums(s)/sum(s),
las = 2,
col = rainbow(10),
ylab = 'Percentage',
main = '% Sentiment Scores for Apple Tweets')
Mr.Baharat, After getting text (tweets) and clean them, How can we rewrite them as a .csv file to get the sentiment analysis of the new data with the clean text. What you did here is that you analysed the raw data with unclean text? Thanks
I wonder the same
Dear Bharatendra ..awesome video thanks for efforts to share knowledge . superb
Thanks for comments!
Hello Professor, first of all, thank you for sharing. I have a question about how to handle a corpus in TXT format.
Thank you sir for a simple tutorial on sentiment analysis and word cloud. Quite helpful for my marketing research class.
Regards.
Thanks for comments!
lettercloud is not work in my r studio ca n you tell me why it showing only spelling of tax , i can't see the result as your's
Recently I too tried and it doesn't show up. Looks like they have discontinued this feature.
@@bkrai kk sir Thank you for replying
You are welcome!
You have explained everything very clearly and it is very helpful for beginners like me. Thank You
Thanks for comments!
Sir I have one doubt. Why have you used : to = "utf-8-mac" in line no.7.
Is it same everytime in all the problems?
@@amanakshansh1021 no, you can just write utf8 instead.
no explanations for the function used. downvoted
Thanks for the feedback!
how can we calculate percentage sentiment of all tweets?
Check this link, I used % here:
ua-cam.com/video/kkMBMxwxTEg/v-deo.html
Excellent. Very well explained. Thanks a lot, Dr.
Most welcome!
this is an amazing tutorial, thank you Dr.!
You're very welcome!
hello sir, i have a doubt, for the first file apple.csv you have done all the pre processing(cleaning of data) ,but for the second time when you are performing the actual sentimental analysis you have not done any pre processing. So you are doing the sentimental analysis without any cleaning in the second file?
Pre processing helps when you are developing wordcloud as you don't want to see thing that are not relevant. However, for sentiment analysis it may not matter if common words like 'the' or 'is' are there or not.
Word clouds should be banned; next to the Pie chart they are one of the worst ways of displaying data.
😃
Nothing to say: pure quality
Thanks!
It is really amazing . Can we know how to get those two files from Twitter ? Is there an R package that enable us to do so? Or how? Thanks.
You can get Twitter data using steps in this link:
ua-cam.com/video/QETCjkQ3CBw/v-deo.html
Honestly while you are typing this reply , am really watching the link you provided me .. much appreciated Bharat. Thanks.
Which algorithm is used for this analysis?
For the sentiment analysis, nrc lexicon is used.
Sir, may i know how did you get "apple2.csv" ?
i.e how did you extract data (apple2.csv) after earnings ?
Here is the link to get tweets in csv file: ua-cam.com/video/QETCjkQ3CBw/v-deo.html
if text form is in different language other than english how can i solve such of text. any video related to NLP bigram,wupalmer, stemming, lemmatization
I've added this to my list.
Many Thans Rai but please explain me how "iconv" function works in simple words.
I didnt understand that function. and Great Work!
It's needed to convert text data to the right format needed when working on a Mac.
what about windows...?
Really excellent video!!
Where can we find dictionaries with keywords annotated with sentiments (e.g. ugly as you mentioned in video)?
Also, how can we add sentiment dictionary for new language? Any video on these topics?
Thank you very much ... :)
Very well explained Bharatendra. I am learning lots of interesting things in R from you.
Thanks for comments!
the code shows error i.e could not find get_nrc_sentiment even i have install syuzhet package can u tell me what is the problem
Make sure to run library line after installing the package.
What if there is an interesting word in the WordCloud. Is it possible to find from which tweet it came from? to localize it basically,
Such a word is likely to be from several tweets. You can know which tweets it came from using the Term Document Matrix.
Mr.Bharat, when i want now to pull new data about apple from Twitter , R returns the tweets dated 2018. My question is: how can i pull tweets ,for example, about the second quarter 2016? Is there any way to specify an exact date range to pull tweets? Regards.
Twitter has some restrictions about how much and how often you can pull data. It only allows you to go back few weeks from current date.
From where u get dataset ??? Source of dataset
Here is the link to get data from Twitter:
ua-cam.com/video/QETCjkQ3CBw/v-deo.html
Can you please make a video on text2vec
I've added it to my list.
Awesome video.. thank you for the great explanation!
Thanks for comments!
Hello,
Is it possible to use the above method for sentimental analysis of tweets in languages other than english?
I've added it to my list.
Thank you very much ... sir we are facing many issues in openNLP package ... request you to kindly help us in getting some insight on the same also ..
I've not used openNLP.
Can I know the issues you are facing? Thanks in advance..
Hello sir,
I am having this issue while using the code for twitter and no idea on how to resolve it.
> setup_twitter_oauth(api_key,api_secret,access_token,access_secret)
[1] "Using direct authentication"
Error in check_twitter_oauth() : OAuth authentication error:
This most likely means that you have incorrectly called setup_twitter_oauth()'
> setup_twitter_oauth(api_key,api_secret,access_token=NULL,access_secret=NULL)
[1] "Using browser based authentication"
Error in init_oauth1.0(self$endpoint, self$app, permission = self$params$permission, :
Forbidden (HTTP 403).
You guidance will be of great help.
For getting Twitter data, try this:
ua-cam.com/video/bxJseJ-aZJk/v-deo.html
hello sir thanks for such a knowledgeable video sir I want to know that which year of data you used in this video and may we use this data for research purposes.please reply thanks again.
It's 2017.
Very informative, thanks a lot. Subscribed the moment after watching this video.
Thanks and welcome!
Awesome videos, examples were very simple and nice, useful materials
Thanks for comments!
Can i get the csv files for twitter data?
Use the link below video.
Hello, I have a question, I would like to ask how could I view an excel file with Chinese characters. After I str my data, it only shows sth like this: U+84EE. I have tried this corpus
For windows use:
corpus
@@bkrai thank you
Hi Bharat; When looking at the data file, we can see that most of columns are N/A. How we can get Longitude and Latitude for example? what is the meaning of replyToSN and replyToSID?
you have mentioned that Apple two datasets are taken from Twitter before and after quarterly earnings report. Could you please let us know what quarter is that? I mean is it after the 1, 2, 3 or fourth quarter the data was taken???
Longitude and Latitude is only available for those Twitter users who agree to share their location. Due to privacy reasons many people do not like to share their location and that's why majority of tweets do not have this info.
replyToSN - reply to Screenname
replyToSID - reply to sender ID
apple2.csv has data for 2nd quarter of 2017.
Too useful , thanks 👌💐💐💐💐
Thanks for comments!
Hello sir,
Don't we have to pre process the text before running get_nrc_sentiment, you are directly passing the comments into that command, but in case of wordcloud you are removing stopwords, urls etc...Could you pls explain why are you doing that?
For sentiment analysis words with no emotions will automatically be ignored. In wordcloud we do not want those words to appear, that are not meaningful.
@@bkrai ok thanks
i have a query, can we do theme tagging for social media data or consumer data ?
Awesome it is! subscribed to you channel :)
Thanks!