Global thesis archiveRDL Thesis
The world's theses in one searchable archive
An open, full-text searchable archive consolidating MSc and PhD theses from around the world into one place. Theses can be assigned a persistent DOI on request.
- Full-text search (title, author, advisor)
- University, department, year and language filters
- Persistent DOI and citation for theses
- Abstract, keywords and full-text PDF
- Advisor and institution pages
RDL ThesisDeep learning for low-resource machine translation
Optimization methods for renewable energy grids
Digital text recognition in historical archives
Microplastics impact on marine ecosystems
Overview
RDL Thesis consolidates MSc and PhD theses from around the world in one open, full-text searchable archive. Search by keyword, author, advisor and title; narrow results with university, department, degree, year and language filters.
Advisor, student and institution pages, plus thesis detail with abstract/keywords/metadata and full-text PDF access. Institutions upload theses one by one or in bulk; a DOI request gives theses a persistent identifier.
See the features up close
Every feature is shown from the real interface with sample data.
Full-text search across millions of theses
Search by title, author, advisor or keyword; results list instantly with title, author, university and year.
Deep learning for low-resource machine translation
Optimization methods for renewable energy grids
Digital text recognition in historical archives
Microplastics impact on marine ecosystems
Narrow the results
Filter by university, department, degree, year and language; reach the thesis you need in seconds.
Refine results
University
Degree
Year
Language
Abstract, keywords and full-text PDF
Thesis detail brings the abstract, keywords, metadata and full-text PDF together, with view and download counts.
Deep learning for low-resource machine translation
Aisha Rahman · Advisor: Prof. Daniel Klein
Abstract
This thesis investigates transfer-learning strategies for neural machine translation in low-resource settings, introducing a multilingual pre-training objective that improves BLEU by 6.4 points across 14 language pairs while reducing data requirements by an order of magnitude.
Deep learning for low-resource machine translation
A. Rahman · Stanford University · 2024
Abstract
- University
- Stanford University
- Department
- Computer Science
- Year
- 2024
- Language
- English
- DOI
- 10.5072/rdl.4821
- Views
- 7,902
Assign a DOI to your thesis, make it permanently citable
Each thesis can be assigned a persistent DOI (Digital Object Identifier) on request; your thesis becomes permanently addressable, citable and indexable, improving citability.
Deep learning for low-resource machine translation
Aisha Rahman · Stanford University · 2024
Digital Object Identifier
10.5072/rdl.thesis.4821
Advisor and institution pages
Discover all theses supervised by an advisor or a university on one page; jump to related work fast.
Prof. Daniel Klein
Stanford University · Computer Science
Supervised theses
From search to full text in three steps
Search, filter, access.
- 1
Search by keyword
Run full-text search by author, advisor, title or keyword.
- 2
Filter results
Narrow by university, department, degree, year and language.
- 3
Access the thesis
Reach the abstract, metadata and full-text PDF from the thesis detail.
Frequently asked questions
Do I need an account to search theses?
No, search and thesis viewing are public.
How does my institution upload theses?
With an institution or admin role, upload one by one or in bulk from the institutional panel.
Can theses get a DOI?
Yes, a DOI request gives theses a persistent identifier and improves citability.
