Comparison of Grammar Dependency and Constituency for Authorship Attribution
| Thesis Type | Bachelor |
| Thesis Status |
Finished
|
| Student | Elisabeth Arzberger |
| Init |
|
| Final |
|
| Start |
|
| Thesis Supervisor | |
| Contact | |
| Research Field |
Various text features have already been tested for their suitability on authorship attribution, including simple features like character and word grams as well as grammatical features like tree distances of constituency trees of texts.
In this thesis, the suitability of dependency parsing in comparison to constituency of grammar is to be analyzed for the field of authorship attribution. The goal is to be able to compare the two approaches using a well-established dataset for authorship attribution.