We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo

Computer Science > Computation and Language

Title: Neural Generation Meets Real People: Towards Emotionally Engaging Mixed-Initiative Conversations

Abstract: We present Chirpy Cardinal, an open-domain dialogue agent, as a research platform for the 2019 Alexa Prize competition. Building an open-domain socialbot that talks to real people is challenging - such a system must meet multiple user expectations such as broad world knowledge, conversational style, and emotional connection. Our socialbot engages users on their terms - prioritizing their interests, feelings and autonomy. As a result, our socialbot provides a responsive, personalized user experience, capable of talking knowledgeably about a wide variety of topics, as well as chatting empathetically about ordinary life. Neural generation plays a key role in achieving these goals, providing the backbone for our conversational and emotional tone. At the end of the competition, Chirpy Cardinal progressed to the finals with an average rating of 3.6/5.0, a median conversation duration of 2 minutes 16 seconds, and a 90th percentile duration of over 12 minutes.
Comments: Published in 3rd Proceedings of Alexa Prize (Alexa Prize 2019)
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
Cite as: arXiv:2008.12348 [cs.CL]
  (or arXiv:2008.12348v2 [cs.CL] for this version)

Submission history

From: Abigail See [view email]
[v1] Thu, 27 Aug 2020 19:37:27 GMT (2146kb,D)
[v2] Sat, 5 Sep 2020 17:14:26 GMT (2135kb,D)

Link back to: arXiv, form interface, contact.