DiScho 🪩 & Social Sciences

Andy Janco, Digital Scholarship Specialist

Digital Scholarship Services

OpenRefine
A powerful free, open source tool for working with messy data

Web Scraping
Consultations on data gathering from the Web that address copyright, good citizenship and API alternatives.

Datasets
Creation and curation of research data from research collections.

Elise Racine & Digit / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/

Combines an image encoder with a large language model allowing images as input.

  • VLMs can recognize typewritten and handwritten text
  • They can retain formatting and layout information
  • Structured data such as tables and forms retain their structure as CSVs or HTML

 

Vision-Language Models

Visual Reasoning (VLRMs)

from pydantic import BaseModel

class Biography(BaseModel):
    id: int
    name: str
    occupation: str
    date_of_birth: date
    place_of_birth: str
    father: str
    mother: str
    education: list[str]
    work_history: list
    children: list
    
PRESIDENCIA DE LA REPUBLICA
COORDINACION EJECUTIVA
ASAMBLEA NACIONAL CONSTITUYENTE
CENTRO DE INFORMACION Y SISTEMAS -5AC-116
IDENTIFICACION C.C. No. 14.976.167 Cali
MESA MUMERO 76001 4007
FECHA DE ELABORACIONADA
ANO
DIA
MUMERO PROPUESTA 07 006
1 2 3 4 5 6
TIPO DE PROPUESTA
1. CONSTITUCION
VIGENTE
INFORMACION PERSONA U ORGANIZACION QUE RESPALDA LA PROPUESTA
UNIVERSIDAD DE SAN BUENAVENTURA
2. ARTICULO
MUEVO

OCR

{
"Mesa Numero": 080016017,
"Fecha de Elaboracion": 901114,
"Numero Propuesta": 0001001,
"Departamento": "Atlantico",
"Municipio": "Baranquilla",
"Proponenta": "Asociación Colombiana de Paramedicos",
...
}

VLMs

# IN SEARCH OF MINIMAL HYPERSURFACES

ANTOINE SONG

A DISSERTATION

PRESENTED TO THE FACULTY

OF PRINCETON UNIVERSITY

IN CANDIDACY FOR THE DEGREE

OF DOCTOR OF PHILOSOPHY

RECOMMENDED FOR ACCEPTANCE

BY THE DEPARTMENT OF

MATHEMATICS

ADVISER: PROFESSOR FERNANDO CODÁ MARQUES

JUNE 2019

Princeton Ph.D. Dissertations and Senior Theses

I would also like to thank Todd Hines from the Firestone library for helping me gather relevant data for this study. 

-- Pui Yan Jane Leong (2015)
To the librarians Joann and Joshua: thank you so much for your quick guidance and direction.

-- Rayleen Hu (2020)

I would also like to thank Bobray Bordelon, my data librarian, for his aid in sourcing data.

 

-- Alex Roose (2024)

Another thanks to Barbara Coffey, the amazing Engineering librarian

-- Ornella Ebongue (2022)

ORF Thesis Acknowledgements