Dear Daniel, all, 

first of all, thanks for interest in that topic.

1) You totally nailed. Additionally, the chat bot should keep track so we can 
answer questions based on the user’s history, e.g.. “When was Barack Obama 
born?” => “Does he have kids?” and resolve he to Barack Obama. 

2) For this GSoC project something more close to ELIZA than to neural nets is 
preferred since neural nets are not easy to control and overwatch. The dialogue 
benchmark is one option to learn the grammar for something ELIZA-like. 

Best regards
Ricardo 
> On 6 Mar 2017, at 14:53, Daniel Obraczka <dan...@obraczka.de> wrote:
> 
> Hello,
> 
> I'm a Computer Science student living in Leipzig currently pursuing my 
> master degree. I'm very interested to work on the chatbot idea. My first 
> warmup task can be found here: 
> https://github.com/dobraczka/DBpedia_chatbot_warmup
> For the second task I wrote a Telegram bot utilizing openQA: 
> https://github.com/dobraczka/OpenQABot
> 
> I also have two questions regarding the DBpedia Chatbot idea:
> 
> 1) The chatbot should build on existing QA systems. Do I understand it 
> correctly, that the role of the chatbot should be on the one hand to 
> give responses where the user is not explicitly asking a question (in 
> which case the QA system cannot be used) and on the other hand ask the 
> user to provide more information if the question is incomplete/ambiguous ?
> 
> 2) The third warm-up task encourages the candidates to read about 
> grammar-based chatbots. Therefore I would assume the aim is to build a 
> closed domain chatbot (similar to e.g. ELIZA), rather than e.g. using 
> neural networks. Is this correct? Or should the dialogue benchmark 
> (mentioned in the project description) be used to train the bot 
> similarly to how e.g. the Ubuntu Corpus[1] was used to train machine 
> learning algorithms?
> 
> Best Regards,
> 
> Daniel Obraczka
> 
> [1] R. Lowe et al.,  The Ubuntu Dialogue Corpus: A Large Dataset for 
> Research in Unstructured Multi-Turn Dialogue Systems, 2015 
> https://arxiv.org/pdf/1506.08909.pdf
> 
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Dbpedia-gsoc mailing list
> Dbpedia-gsoc@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/dbpedia-gsoc

-- 
Ricardo Usbeck, M.Sc.
Augustusplatz 10
Room P905
04109 Leipzig

http://aksw.org/RicardoUsbeck

------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Dbpedia-gsoc mailing list
Dbpedia-gsoc@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/dbpedia-gsoc

Reply via email to