Hacker News new | past | comments | ask | show | jobs | submit login
How I Used Deep Learning to Train a Chatbot to Talk Like Me (adeshpande3.github.io)
220 points by adeshpandedsfd on Aug 10, 2017 | hide | past | favorite | 33 comments



It's always nice to read the details of people building real applications, even if it's not a homerun.

BTW: just blurring people's names using a gaussian blur operator is not enough, if your radius is too small. From the screenshots of chats, you can tell names like "Arvind Sankar", "Manu Saravanan", etc.


Also, although difficult, Gaussian blurs can be in certain cases reversed (even though this one is so obvious there was no point in blurring the names). I prefer a fat black box when censoring.


If you want a more robust parser I made scripts to export Messenger and Hangouts chat logs to dataframes for a previous project: https://github.com/MasterScrat/ChatShape

It will give you rows with : [timestamp, interlocutorName, messageSenderName, text]


Thanks! This is definitely useful


This reminds me of this year's xyzzy award-winning interactive fiction game, The Mary Jane of Tomorrow. https://emshort.blog/2016/06/05/the-mary-jane-of-tomorrow/ (It won Best Single Puzzle and Best NPC.)



Just watched the episode today LOL Insane stuff


Great explanation! I also worked on a similar project (https://medium.com/towards-data-science/personality-for-your...).

Also feel free to check my Conversation Analyzer (https://github.com/5agado/conversation-analyzer), which includes scraper and parser for Facebook Messanger conversations.


Nice writeup! Definitely seems to be a lot more realistic than my bot LOL The parser is really helpful too


Is it possible to make it spew out random personal information like addresses and phone numbers if you give it the right input? That's usually what happens when you train a RNN and it overfits.


Yeah you're right it is a possibility. Overfitting was definitely a problem in this project. I think just a larger dataset and maybe adding some regularizers would have helped.

Luckily I dont think I have too much compromising information in my dataset LOL


Doesn't seem like it's really any better than eliza or the emacs doctor or whatever stuff didn't use the buzzword tech of the year.


Ha! Just yesterday I had the narcisstic idea of gifting someone I've been chatting regularly with for years a chatbot that answers his messages like I would. This is just what I needed, thanks!


Awesome! Let me know if you get good results or find a way to improve my Seq2Seq model


I have this idea pretty regularly. Human interactions are quite often annoying :).


Really cool project! Curious to see what type of changes you could make to get more realistic results


while(0) {cout << 'hurr. durr.\n'; sleep(1000);}


So exit immediately?


Hi. Welcome to chatbot/n.


[flagged]


I think we don't need this off-topic flamebait here. Please post civilly and substantively or not at all.

https://news.ycombinator.com/newsguidelines.html


Funny how someone without anything remotely technical in their comment history feels like they can speak for "the majority of tech people".


ad hominem super sleuth


I don't see anything bad that he said. Also, I've met a decent amount of people who come across as very "bro" who are actually really nice considerate people so being kinda bro isn't inherently bad (unless you define "bro" as essentially just being an ahole, which doesn't seem to apply here)


So.. I went reading the article, expecting something really ridiculously bro. The article was well written, good references, but yes, uses a few silly conversation samples (for which there were no sexist, racist or homophobic references).


Is it possible that training a bot to speak like that (silly/casual) is more challenging than to emulate a very grammatically correct and stiff style? As such, maybe the choice of conversations was intentional?


My comment was directed towards the OP, who was harsh and didn't provide a proper argument. As I said, it was a nice article and I wanted to encourage others to read it, despite that (flagged) comment.


Yes, I was writing in support of your comment.


Being bro is synonymous with being male to some people.

It is, of course, still an insult to them.


I see it more often used to mock a male mono-culture and the pitfalls that derive from it (ex: brogrammers). A bit like the term douchebag has become a synonym of 'jerk'.


The author is probably 20 years old (give or take) Undergrad student.

What is wrong with you people? Always have to put others in their "stereotypical silos" and stupid them they are the racists and homophobes ?!


Author seems to be into fantasy sports and must have a lot of group threads on that topic.


Yeah LOL i think maybe limiting the dataset to just 1:1 conversations would be better than including group chats as well


Ideally most of them won't care what other people think about them. That's not something to be spending much thinking power on anyway.

There's nothing wrong with making tech that can communicate to a larger audience than "people in tech".




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: