Master’s Thesis Presentation • Machine Learning — Dialog Response Generation Using Adversarially Learned Latent Bag-of-WordsExport this event to calendar

Tuesday, August 18, 2020 2:30 PM EDT

Please note: This master’s thesis presentation will be given online.

Kashif Khan, Master’s candidate
David R. Cheriton School of Computer Science

Dialog response generation is the task of generating response utterance given a query utterance. Apart from generating relevant and coherent responses, one would like the dialog generation model to generate diverse and informative sentences.

In this work, we propose and explore a novel multi-stage dialog response generation approach. In the first stage of our proposed multi-stage approach, we construct a variational latent space on the bag-of-words representation of the query and response utterances. In the second stage, transformation from query latent code to response latent code is learned using an adversarial process. The final stage involves fine-tuning a pretrained transformer based model called text-to-text transfer (T5) (Raffel et al., 2019) using a novel training regimen to generate the response utterances by conditioning on the query utterance and the response word learned in the previous stage.

We evaluate our proposed approach on two popular dialog datasets. We show using quantitative metrics that our proposed approach outperforms the state of the art approaches in generating diverse responses while still performing competitively on other quantitative metrics.


To join this master’s thesis presentation on Zoom, please go to https://us02web.zoom.us/j/83677626723?pwd=WElYWWc2ZWQ2eE1VR3BsWERScFczZz09

Meeting ID: 836 7762 6723
Passcode: 047655

Location 
Online presentation
200 University Avenue West

Waterloo, ON N2L 3G1
Canada

S M T W T F S
28
29
30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
1
  1. 2024 (100)
    1. April (23)
    2. March (27)
    3. February (25)
    4. January (25)
  2. 2023 (296)
    1. December (20)
    2. November (28)
    3. October (15)
    4. September (25)
    5. August (30)
    6. July (30)
    7. June (22)
    8. May (23)
    9. April (32)
    10. March (31)
    11. February (18)
    12. January (22)
  3. 2022 (245)
  4. 2021 (210)
  5. 2020 (217)
  6. 2019 (255)
  7. 2018 (217)
  8. 2017 (36)
  9. 2016 (21)
  10. 2015 (36)
  11. 2014 (33)
  12. 2013 (23)
  13. 2012 (4)
  14. 2011 (1)
  15. 2010 (1)
  16. 2009 (1)
  17. 2008 (1)