SLIDE 1
Understanding User Interactions with Podcast Recommendations Delivered Via Voice
1
Funders:
Lo Longqi Yang Deborah Estrin Michael Sobolev Christina Tsangouri
Emerging Voice Interfaces for Content Consumption
2
Understanding User Interactions with Podcast Recommendations - - PowerPoint PPT Presentation
Understanding User Interactions with Podcast Recommendations - - PowerPoint PPT Presentation
Understanding User Interactions with Podcast Recommendations Delivered Via Voice Lo Longqi Yang Michael Sobolev Christina Tsangouri Deborah Estrin Funders: 1 Emerging Voice Interfaces for Content Consumption 2 Diverse Content Delivered
SLIDE 2
SLIDE 3
Diverse Content Delivered Through Voice Interfaces
3
SLIDE 4
4
Visual Interface Voice Interface
Hi Hi, Al Alice, e, here ere are re the e ep episodes es rec recommen ended ed to you
- u – nud
nudge from the ted radio hour ur, , … Pl Play y the e ev evolutio ion of artif ific icia ial in intel ellig igen ence e from the e fres esh air ir
Episode #1 Episode #2 Episode #3 Episode #4 …
SLIDE 5
5
- Users may spend longer time considering
each recommendation.
- Users may explore recommendation lists less.
- Users may less likely choose items ranked
lower in recommendation lists.
What we may expect
How much worse it is? Does it matter?
SLIDE 6
6
This Work
A A be between-su subject ect r random
- mized
ed c con
- ntrol
- l st
study y co comparing Voice and Visual interfaces Ef Effici cien ency Ch Choic ice Ex Exploration
SLIDE 7
7
Study Design
Gr Group p B Gr Group p A Independent variable (Factor): In Interface Same list of podcast recommendations
SLIDE 8
8
Study Design
St Step 1:
- Br
Browse a long list of episodes (1-N).
- Making
g a co committed ch choi
- ice:
ce: typing in the index number & listen to the episode for at least 5 minutes
Independent variable (Factor): In Interface Same list of podcast recommendations Gr Group p B Gr Group p A
SLIDE 9
9
Study Design
St Step 1:
- Br
Browse a long list of episodes (1-N).
- Making
g a co committed ch choi
- ice:
ce: typing in the index number & listen to the episode for at least 5 minutes St Step 1:
- Lis
Listen n to a long list of episodes (1-N).
- Making
g a co committed ch choi
- ice:
ce: typing in the index number & listen to the episode for at least 5 minutes
Independent variable (Factor): In Interface Same list of podcast recommendations Gr Group p B Gr Group p A
SLIDE 10
10
Study Design
Same list of podcast recommendations
St Step 1:
- Br
Browse a long list of episodes (1-N).
- Making
g a co committed ch choi
- ice:
ce: typing in the index number & listen to the episode for at least 5 minutes St Step 1:
- Lis
Listen n to a long list of episodes (1-N).
- Making
g a co committed ch choi
- ice:
ce: typing in the index number & listen to the episode for at least 5 minutes St Step 2:
- Listen to the episode for at least 5
minutes St Step 2:
- Listen to the episode for at least 5
minutes
Independent variable (Factor): In Interface Gr Group p B Gr Group p A
SLIDE 11
11
Study Design
St Step 1:
- Br
Browse a long list of episodes (1-N).
- Making
g a co committed ch choi
- ice:
ce: typing in the index number & listen to the episode for at least 5 minutes St Step 1:
- Lis
Listen n to a long list of episodes (1-N).
- Making
g a co committed ch choi
- ice:
ce: typing in the index number & listen to the episode for at least 5 minutes St Step 2:
- Listen to the episode for at least 5
minutes St Step 2:
- Listen to the episode for at least 5
minutes St Step 3:
- A user satisfaction survey
St Step 3:
- A user satisfaction survey
Independent variable (Factor): In Interface Same list of podcast recommendations Gr Group p B Gr Group p A
SLIDE 12
12
Study Design
Independent variable (Factor): In Interface 100 50 50 IT ITunes trendi ding g po podc dcasts Gr Group p B Gr Group p A
SLIDE 13
13
Efficiency
The number of recommendations considered per minute Vo Voice Vi Visual 54 6
SLIDE 14
14
Exploration
The maximum index number of the recommendations considered Vo Voice Vi Visual 47.8* 15.3
SLIDE 15
15
Choice: Lower-ranked Items
The index number of the chosen episode Vo Voice Vi Visual 47.8 7.3
SLIDE 16
16
Choice: Lower-ranked Items
preferred item (choice) stop
SLIDE 17
17
Main Implications
1/ 1/3 exploration 1/ 1/6 chosen episode rank 1/ 1/9 efficiency
SLIDE 18
18
1/ 1/3 exploration 1/ 1/6 chosen episode rank 1/ 1/9 efficiency Better navigation techniques
Main Implications
SLIDE 19
19
Better navigation techniques Adaptive and diverse recommendations at top ranks 1/ 1/3 exploration 1/ 1/6 chosen episode rank 1/ 1/9 efficiency
Main Implications
SLIDE 20
20
Lo Longqi Yang
Ph.D. candidate Computer Science, Cornell Tech, Cornell University Email: ylongqi@cs.cornell.edu Web: bit.ly/longqi Twitter: @ylongqi Connected Experiences Lab http://cx.jacobs.cornell.edu/ Small Data Lab http://smalldata.io/ Funders: