Lisa Hoek

Who I Am

Hi, I'm Lisa Hoek, a passionate and experienced Data Specialist working at the City Archives of Amsterdam, where I transform and optimize metadata to make their collections more accessible. I used to travel around Europe in my campervan, but I traded the digital nomad lifestyle for a home base in the Netherlands.

I have strong skills in Python, using packages like pandas, matplotlib, and rdflib. My focus areas are machine learning, AI, and Natural Language Processing (NLP), and I enjoy working with both text and image data.

One of my past projects is transcribing 20th century Surinamese and Curaçaoan death certificates using Transkribus and GPT. I also work a lot with regular expressions, Named Entity Recognition, and Linked Open Data.

Besides that, I specialise in prompt engineering, using Large Language Models (LLMs) to build smart applications. My Master's degree in Data Science helps me pick up new tools and challenges quickly.

Expertise

Prompt Engineering (LLMs)

Natural Language Processing

Automatic Text Recognition

Machine Learning

Data Analysis

Computer Vision

Linked Open Data

Portfolio

View more

Smart Rule implementation using SpaCy

Smart Rule implementation using SpaCy

View more

Transcribing Surinamese death certificates using Transkribus and GPT

Transcribing 20th century Surinamese death certificates using Transkribus and GPT

View more

GPT prompt revision for kids' magic pen pal

GPT Prompt revision for kids' magic pen pal

View more

Detecting age discrimination in job advertisements using GPT and BERT

Detecting Age Discrimination in Job Advertisements using GPT and BERT

View more

Transcribing Curaçaoan death certificates using Transkribus and GPT

Transcribing 19-20th century Curaçaoan death certificates using Transkribus and GPT

View more

Extracting entities from handwritten civil records using HTR and RegExes

Extracting Entities from Handwritten Civil Records using HTR and RegExes (MSc thesis)

View more

Pull-factor regular expressions in job advertisements

Pull-Factor Regular Expressions in Job Advertisements

View more

Named entities in job advertisements

Named Entities in Job Advertisements
(Internship Data Science)

View paper

Analysing cat and dog popularity based on attributes in their photos

Analysing cat and dog popularity based on
attributes in their photos

View thesis

Building a web classification system using DMOZ

Building a web classification system using DMOZ (BSc thesis)

What colleagues say

Björn Quanjer

Radboud University

Lisa Hoek is a highly recommended professional known for her impressive growth and exceptional qualities as a researcher. Starting with strong programming skills from her computer science background, she quickly mastered historical sources through her eagerness to learn and systematic approach. Her work led to efficient digitization of historical documents, accelerating and enriching related research. She excels in open communication, independent problem-solving, and innovative thinking, making her a valuable asset to any team or project.

Anna Pillar

Textmetrics

I am working together with Lisa for a bit over 2 years by now and it has been such a pleasant experience. Lisa is responsive and reliable. Communication with her is super easy and she keeps me in the loop. She reaches out to me regularly to make sure we are on the same track with decisions that have to be made during a project. From a technical point of view, I have full faith in her abilities and she delivers great work.

Erik Tjong Kim Sang

eScience Center

I worked with Lisa in the 2024 Radboud University project on automatically transcribing nineteenth-century hand-written colonial demographic documents. Lisa proved to be very knowledgeable on the task. She developed several tests to decide which parts of the task could accurately be done automatically. I hope we will have a chance to collaborate with Lisa in the future!