Doctoral Theses in France (1985-2025): A Linked Dataset of PhDs, Academic Networks, and Institutions
This paper presents a comprehensive, linked dataset of French doctoral theses from 1985 to 2025, constructed by aggregating and enriching national metadata to enable detailed analysis of academic careers, networks, and institutional evolution.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine a massive, dusty library in France that holds the stories of every single PhD thesis defended there since 1985. For decades, these stories were scattered across different shelves, written in different handwriting, and sometimes missing key pages like "who was on the judging panel" or "what year the professor was born."
This paper is the story of a team of librarians (the authors) who decided to clean up this library, organize it, and turn it into a super-powered, searchable digital map.
Here is the breakdown of what they did, using some everyday analogies:
1. The Problem: A Messy Attic
Before this project, if you wanted to study how PhDs work in France, you'd have to dig through a chaotic attic. You might find a thesis title, but the list of judges (the "jury") might be missing, or the names of the professors might be spelled differently in different records. It was like trying to solve a puzzle where half the pieces are from a different box.
2. The Solution: The "Master Key" System
The authors built a new dataset by taking the official French government list of PhDs (called Thèses.fr) and giving it a serious makeover. Think of this process as three main steps:
- Step 1: The Detective Work (Cleaning): They found names that were mixed up (like "First Last" vs. "Last First") and fixed them. They also found "ghost" IDs—numbers that were supposed to link to a person but didn't actually exist. They manually fixed 168 of these broken links so everything connects properly.
- Step 2: The Crystal Ball (Guessing): The original list didn't always say if a professor was a man or a woman. So, the authors used a smart computer tool (like a very advanced name-guessing app) to infer gender based on first names. This filled in a huge gap, turning thousands of "unknowns" into known data points.
- Step 3: The Super-Connector (Linking): They didn't just stop at the thesis list. They grabbed "ID cards" from other massive French databases (like IdRef, TEL, and SUDOC).
- Analogy: Imagine you have a phone number for a person. This project added their home address, their birth date, and their social security number to that same phone number. Now, you can instantly see their whole life story, not just their name.
3. What's Inside the Box?
The final result is a giant spreadsheet (a dataset) that covers 40 years of French academic history. It's organized into six main "folders":
- The Thesis: The title, the date, and the topic (like "Medicine" or "History").
- The Student: Who defended the thesis?
- The Bosses: Who supervised the student?
- The Judges: Who sat on the panel to say "Pass" or "Fail"?
- The Schools: Which universities and research labs were involved?
- The Content: The abstract and keywords in many different languages.
4. Why Does This Matter? (The "Aha!" Moments)
This isn't just a list of names; it's a time machine for studying how science and academia work.
- Mapping the "Who Knows Who" Network: Because the dataset lists every judge and every supervisor, researchers can draw a giant map of who knows whom. They can see if certain professors always work together, or if a specific university is a "hub" that connects to everyone else.
- Analogy: It's like mapping the entire cast of a TV show to see who is friends with whom, and how the "cool kids" (central researchers) are connected to the new students.
- Tracking Time: You can see how things changed over 40 years. For example, the paper notes that in the past, thesis juries were small. Now, they are much bigger. This dataset proves that change with hard numbers.
- The "Missing Piece" of History: The authors explain that before 2006, French universities didn't have formal "Doctoral Schools." So, if you look at data from 1990, you won't see school names. This isn't a mistake in the data; it's a historical fact that the dataset preserves perfectly.
5. The Big Picture
The authors made this data free for anyone to use. They want sociologists, historians, and data scientists to use it to answer big questions like:
- "Are women getting more seats on PhD judging panels over time?"
- "How do research teams form across different countries?"
- "How has the language of science changed?"
In short: This paper takes a messy, fragmented history of French PhDs, polishes it, connects all the dots, and hands you a shiny, interactive map so you can explore the hidden patterns of how knowledge is created and shared.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.