Encouraging and enabling conciseness in text production
Event details
Date | 01.06.2018 |
Hour | 14:00 › 16:00 |
Speaker | Kristina Gligoric |
Location | |
Category | Conferences - Seminars |
EDIC candidacy exam
Exam president: Prof. Pierre Dillenbourg
Thesis advisor: Prof. Robert West
Co-examiner: Prof. Karl Aberer
Abstract
This proposal describes the interplay between constraints and text content production on online social media. Our research goal is providing a better understanding of the extent to which users can benefit from producing concise content and developing the ways of supporting them, thus enabling conciseness in text production. In this proposal, three related works and their relation to this line of research are discussed. Firstly, a recent background study by Tan et al., revealing that wording has significant effects on message propagation and examining the features that improve it on Twitter is discussed. Secondly, work by Bernstein et al. introducing a tool called Soylent is described. It demonstrates the feasibility of engaging crowds to assist users in shortening text such as blogs or technical papers. Finally, an attention-based neural model for abstractive sentence summarization developed by Rush et al. is considered in order to examine whether this beneficial process can be automated.
Background papers
The effect of wording on message propagation: Topic- and author-controlled natural experiments on Twitter, by Tan, C., et al.
Soylent: a word processor with a crowd inside, by Bernstein M., et al.
A Neural Attention Model for Abstractive Sentence Summarization, by Rush, A., et al.
Exam president: Prof. Pierre Dillenbourg
Thesis advisor: Prof. Robert West
Co-examiner: Prof. Karl Aberer
Abstract
This proposal describes the interplay between constraints and text content production on online social media. Our research goal is providing a better understanding of the extent to which users can benefit from producing concise content and developing the ways of supporting them, thus enabling conciseness in text production. In this proposal, three related works and their relation to this line of research are discussed. Firstly, a recent background study by Tan et al., revealing that wording has significant effects on message propagation and examining the features that improve it on Twitter is discussed. Secondly, work by Bernstein et al. introducing a tool called Soylent is described. It demonstrates the feasibility of engaging crowds to assist users in shortening text such as blogs or technical papers. Finally, an attention-based neural model for abstractive sentence summarization developed by Rush et al. is considered in order to examine whether this beneficial process can be automated.
Background papers
The effect of wording on message propagation: Topic- and author-controlled natural experiments on Twitter, by Tan, C., et al.
Soylent: a word processor with a crowd inside, by Bernstein M., et al.
A Neural Attention Model for Abstractive Sentence Summarization, by Rush, A., et al.
Practical information
- General public
- Free
Contact
- EDIC - [email protected]