One Last Project
Suppose we’d like to understand the bonds between pairs of people in a book!
- How can we find characters in a story?
- How can we infer connections between characters?
- How can we draw a graph?
Friends?
Given the text from a novel, how can we infer interaction or connections between characters? Discuss this question, and write down your ideas.
________________________________________
________________________________________
________________________________________
________________________________________
We Are Not the First
![]()
Starting From the Goal
The social network graph will have…
Vertices:
Edges:
Edges
We could consider every pair of people and check every paragraph for their presence.
OR, we could
Edges
"I'm glad," said Anne quickly. "I couldn't really feel proud of keeping head of little boys and girls of just nine or ten. I got up yesterday spelling 'ebullition.' Josie Pye was head and, mind you, she peeped in her book. Mr. Phillips didn't see her—he was looking at Prissy Andrews—but I did. I just swept her a look of freezing scorn and she got as red as a beet and spelled it wrong after all."
- What names appear?
- What pairs should be tallied?
- General observations:
Moving to the Middle
Given [[AS, JP, MP], [MP, PA], [AS, JP]],
- Could you create
[AS, JP, MP, PA]?
- and
{(AS,JP):2, (AS,MP):1, (JP,MP):1, (MP,PA):1}?
Demo
In the PL activity, load socnet_nb.py.
