Clean Up and Update Single Exchange Dialogs

[Nirant]

The Single Exchange Dialogs section is ambiguous, too broad and out of date. Here is how you can help us improve this:

1. Removing links which you think do not fit in the section. Don't worry about damaging this repository, we can have a discussion on the PR you raise.
2. Consider adding 2-3 code examples and datasets
3. Consider adding 2-3 examples from slot filling (sequence mining for text is welcome too) and other approaches in chatbots

[Anurag Sharma]

Hey @NirantK , in code examples can I add how one can use CountVectorizer and Tf-idfVectorizer? And how pre-trained vectors can be used in NLP?

[Nirant]

Hey @anu0012 , if the vectorization methods specific to Dialogs and chatbots - feel free to add them to that section. I guess they are not.

If you are asking if you can add them in general, consider adding them to the tutorial section - if these are not already covered.

[Anurag Sharma]

https://github.com/anu0012/Predict_the_happiness_challenge/blob/master/notebook.ipynb

In this notebook, I have used Tf-IDF Vectorizer. I used several concepts like text-cleaning, lemmatization, stemming etc. in this script. Can I add this?

[Nirant]

No, @anu0012 that does not meet our requirements just yet. Please refer the tutorials section to get an estimate of the quality needed to be included here.

I am sure you can polish it to make it awesome and help the community in the process!

[Dhruv Apte]

@NirantK, I first checked the links here.

1. the RNNLM toolkit link is broken.
2. The other papers have working links
   New stuff worth adding

• SPMF , a Java library for pattern mining
• A Sequential pattern mining tutorial and a 'hands-on' thingy
• This code repo is dual LSTM encoder for dialog response generation from the Ubuntu corpus.
  Anything I am missing out or mistaking for something?

[Nirant]

Thanks for looking into this! Single Exchange Dialogs are not traditional Sequence Mining Problems. The traditional Sequence Mining problems include techniques to answer questions like if a customer buys a car X, will he buy an insurance within first 2 weeks, 2 months? And at what price point? So the first 2 new links are not relevant to NLP.

…

--- Please open broken RNNLM toolkit as a separate issue - I'll look into it.

--- check https://dialogflow.com as an example of service offering dialogue/HCI systems here is the link to the ubuntu corpus paper: https://arxiv.org/abs/1506.08909 check what are the most interesting/most cited works which have cited the ubuntu corpus above? Hope this helps you get kick started! Thanks again and sincere apologies if I've been too harsh.

On 17 January 2018 at 22:05, Dhruv Apte ***@***.***> wrote: @NirantK <https://github.com/nirantk>, I first checked the links here. 1. the RNNLM toolkit link is broken. 2. The other papers have working links New stuff worth adding - SPMF <http://www.philippe-fournier-viger.com/spmf/> , a Java library for pattern mining - A Sequential pattern mining <http://data-mining.philippe-fournier-viger.com/introduction-sequential-pattern-mining/> tutorial and a 'hands-on' thingy <http://data-mining.philippe-fournier-viger.com/tutorial-how-to-discover-hidden-patterns-in-text-documents/> - This code repo <https://github.com/dennybritz/chatbot-retrieval/> is dual LSTM encoder for dialog response generation from the Ubuntu corpus. Anything I am missing out or mistaking for something? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#96 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ADGaPUiDWUetrVc0HNpfJSZwuklPM6yIks5tLiFVgaJpZM4RXbAb> .

[Dhruv Apte]

I understand. 😄 . I mistook it for something. No problem, I will open up the needed issue and look more in the DialogCI and ubuntu corpus thing

[Anurag Sharma]

https://www.tidytextmining.com

I think this can be added in reading section. What do you think @NirantK ?

[Nirant]

@anu0012 good find. Since this is an entire book and not a one-off tutorial, let's create a new section under tutorials Books and add there.

This becomes our excuse to make some progress on #5 as well.

[Nirant]

Thank you @the-ethan-hunt.

I have fixed the broken link and closed #105.

As a quick note, Dialogflow is a tool for making Human-Computer Interaction systems (or HCI). In layman words, it is a tool for making chatbots.

[Nirant]

Hey @the-ethan-hunt, do consider continue contributing to awesome-nlp. Take a look at this issue if you'd like :)

[Dhruv Apte]

Sure @NirantK ! But is there any other issue I might possibly work on? 😅

[Nirant]

Sure @the-ethan-hunt.

Thanks for adding Korean from #98 but did not make enough progress on Chinese, Japanese or any European languages for that matter. It'd be awesome if we'd take that issue to its due conclusion.

It saves a lot of time for the community to have all of the best tools for a particular language in one place.

[Nirant]

Hey @anu0012, are you still interested in working on this? We could really appreciate a hand here :)

[Anurag Sharma]

Sure @NirantK. In the second point which you mentioned what type of code examples and dataset can be added?

[Nirant]

@anu0012 Chatbots, virtual assistants and any other popular form of conversational interfaces is a good starting point.

E.g. there is some work on chatbots from Microsoft and Facebook both, check for what datasets they've used and if we can mention them here. Similarly, there is some work on intent detection etc, maybe look if that is relevant?

If at the end of all of this search, we are still unsatisifed with the quality and breadth of coverage, maybe we can merge this section with Conversational Q&A which has similar technical challenges imho. I'd be mostly going by your (and community's) recommendation and findings on the same.