
[08] He He - Sequential Decisions and Predictions in NLP
The Thesis Review
00:00
How to Model the Knowledge Graph of Two Different Agents
In our setting, each agent has access to the self-friends with different attributes. And so dialogue could go like one person says, I have friends working at NYU, and then they went to Google. The other person can say, oh, I also have friends who are at Google and they studied computer science. They basically try to match attributes of their friends to figure out who's the mutual one.
Transcript
Play full episode