Representation learning methods for computational social science
Abstract/Contents
- Abstract
- The combination of machine learning and massive social datasets has the potential to revolutionize our ability to predict and understand human behavior. However, machine learning on large social datasets is difficult because these datasets tend to be multimodal, dynamic, and involve graph-structured relationships (e.g., social networks between users)---while traditional machine learning tools are largely designed for static datasets comprised of simple Euclidean vectors or grids. In this dissertation, we address this challenge and describe new methods for machine learning on massive, graph-structured social datasets. The technical focus of this dissertation is on techniques for graph representation learning, i.e., representation learning with graph-structured or relational data. We show how we can use graph representation learning to model social systems in a data-driven way---by learning social representations or embeddings directly from raw data, rather than relying on painstaking and brittle feature engineering. We introduce techniques to learn social representations of individuals, communities, words, and generic social media content, and we highlight the utility of such embeddings in a large number of applications, including using social representations to uncover new laws of language evolution and to power a social recommender system serving millions of users. The key contributions of this dissertation are twofold: First, we describe new representation learning algorithms that are specifically designed to accommodate the relational complexities of social data. Second, we show---through numerous applications---how learned social representations can facilitate new kinds of social science and new kinds of social applications.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2018; ©2018 |
| Publication date | 2018; 2018 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Hamilton, William L |
|---|---|
| Degree supervisor | Jurafsky, Dan, 1962- |
| Thesis advisor | Jurafsky, Dan, 1962- |
| Thesis advisor | Leskovec, Jurij |
| Thesis advisor | Ré, Christopher |
| Degree committee member | Leskovec, Jurij |
| Degree committee member | Ré, Christopher |
| Associated with | Stanford University, Computer Science Department. |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | William L. Hamilton. |
|---|---|
| Note | Submitted to the Computer Science Department. |
| Thesis | Thesis Ph.D. Stanford University 2018. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2018 by William Leif Hamilton
- License
- This work is licensed under a Creative Commons Attribution 3.0 Unported license (CC BY).
Also listed in
Loading usage metrics...