Day 19 of #100daysofnetworks
Artificial Life, Bipartite Projection, and Simple NLP
Earlier this year, I was reading a book called The Ascent of Information. The entire book is a whirlwind of insights as well as a page turner. I couldn’t put it down. Later in the book, the field of Artificial Life is discussed, and I became intrigued the more I read. If you haven’t read this book, it is one of my favorites.
Last week, I created an arXiv data collector, so that I can quickly research any topic of interest, and analyze the collaboration networks that exist in the topic ecosystem.
Today, I used that arXiv collector to investigate Artificial Life. I think I have always been interested in this field. Artificial Intelligence gets all of the attention, but there are people working on Artificial Life research as well. It is a very cool field of study.
Code is Available
There are two Jupyter notebooks this week:
Bipartite Projection Correction
In one of the steps used during this blog series, I did not quite understand what was needed for one parameter of a function, so I did it wrong. It worked, which confused me. It wasn’t a breaking error, but it essentially created a network of authors as well as a network of titles. Today’s code has the correction, for how to properly do Bipartite Projection, to build these collaboration networks.
It looks like this. You are supposed to provide the nodes that you want to appear in the projected graph. I was giving it every node, which isn’t the way it should be used. Good to know!
I also suspect that this projection approach is a better approach than how I mapped out the Alice in Wonderland network in my book, so I will be playing with that idea in later days, to see if I can improve the network.
What’s in the Artificial Life Network?
I have a lot to cover and limited space on Substack, so will go fast. Play with the code to learn more, and to do your own research on your own topics of interest!
The arXiv search engine was returning too much junk data about Artificial Intelligence, so I used a filter today to only keep articles that explicitly mention “artificial life” in the summaries. This is the first simple NLP trick, using NLP to cleanup a graph.
After the filter, a very small network remained.
Graph with 307 nodes and 2026 edges
This is easily small enough to visualize.
I think this is beautiful. It shows that this is a small network. I think of it as quiet. Large networks are much more complex. This has all of the interesting pieces I would hope to find:
Isolate nodes (people who wrote a paper alone and only one paper)
Connected components (small groups that worked on a paper)
Dense ecosystem (people who have collaborated on multiple papers)
This is enough to also be able to identify some of the most active authors in the collaboration network.
If someone wanted to enter the field of Artificial Life, these might be people worth getting to know. They have written papers, and their network positioning shows their influence.
Some of the ego networks are very interesting and reveal something about papers. There are some papers that have dozens of authors.
In this image, it shows that Kenneth likely collaborated on at least three papers. One with Jay, one with the group on the top, and one with the group on the bottom right. But that supercluster is unusual, and now I am noticing them more on arXiv.
If I look at the core of the network, I can see that there are three of these clusters.
What Articles Are There?
Today, I want the focus to be on usability of these networks. I want to show what papers I’ve found, more than I want to talk about networks. I want this to be useful, not just neat.
If you look at the code, you can see that this is not a fast-moving field in terms of number of papers written per month. It’s not like Artificial Intelligence, where it’s impossible to keep up with the research.
After applying the filter to remove non-ALife papers, the maximum number of papers per month is four. The topic is also written about in different categories. Read about arXiv’s category taxonomy.
The top three categories are:
Computer Science: (Neural and Evolutionary Computing)
Computer Science: (Artificial Intelligence)
Nonlinear Sciences: (Adaptation and Self-Organizing Systems)
Good to know! Now, to top off this article, I’ll link to as many papers as I have space, the ones that look most interesting to me today.
Artificial Life Papers by Category
Category: cs.NE
An artificial life approach to studying niche differentiation in soundscape ecology
Artificial Life in Game Mods for Intuitive Evolution Education
Increasing Behavioral Complexity for Evolved Virtual Creatures with the ESP Method
JohnnyVon: Self-Replicating Automata in Continuous Two-Dimensional Space
Matchmaker, Matchmaker, Make Me a Match: Geometric, Variational, and Evolutionary Implications of Criteria for Tag Affinity
Modelling SARS-CoV-2 coevolution with genetic algorithms
Open Questions in Creating Safe Open-ended AI: Tensions Between Control and Creativity
Qualities, challenges and future of genetic algorithms: a literature review
Role of Morphogenetic Competency on Evolution
Safe Mutations for Deep and Recurrent Neural Networks through Output Gradients
Self-Organizing Intelligent Matter: A blueprint for an AI generating algorithm
Self-Replicating Machines in Continuous Space with Virtual Physics
SignalGP-Lite: Event Driven Genetic Programming Library for Large-Scale Artificial Life Applications
Sooner than Expected: Hitting the Wall of Complexity in Evolution
The Importance of Open-Endedness (for the Sake of Open-Endedness)
The Self-Organizing Symbiotic Agent
The Surprising Creativity of Digital Evolution: A Collection of Anecdotes from the Evolutionary Computation and Artificial Life Research Communities
Towards Large-Scale Simulations of Open-Ended Evolution in Continuous Cellular Automata
Towards a Framework for Observing Artificial Evolutionary Systems
WebAL-1: Workshop on Artificial Life and the Web 2014 Proceedings
Category: cs.AI
An Inductive Formalization of Self Reproduction in Dynamical Hierarchies
Chemlambda, universality and self-multiplication
Emotional Responses in Artificial Agent-Based Systems: Reflexivity and Adaptation in Artificial Life
Evolved Open-Endedness in Cultural Evolution: A New Dimension in Open-Ended Evolution Research
From Alife Agents to a Kingdom of N Queens
Hybrid Life: Integrating Biological, Artificial, and Cognitive Systems
Intelligence as information processing: brains, swarms, and computers
Motility at the origin of life: Its characterization and a model
Perspective: Purposeful Failure in Artificial Life and Artificial Intelligence
Quantifying Natural and Artificial Intelligence in Robots and Natural Systems with an Algorithmic Behavioural Test
Category: nlin.AO
Introduction to Random Boolean Networks
Network Complexity of Foodwebs
The differences between natural and artificial life. Towards a definition of life
Category: adap-org
Evolutionary Learning in the 2D Artificial Life System "Avida"
Propagation of Information in Populations of Self-Replicating Code
Category: physics.gen-ph
Evolution in the Multiverse
The Evolution and Development of the Universe
Category: nlin.CG
Classification of Complex Systems Based on Transients
Evolving Structures in Complex Systems
Lenia - Biology of Artificial Life
Lenia and Expanded Universe
Category: q-bio.PE
Evolution of Complexity
Category: cs.RO
Sustainably Grown: The Underdog Robots of the Future
Category: cs.IT
JIDT: An information-theoretic toolkit for studying the dynamics of complex systems
Mechanical generation of networks with surplus complexity
Towards a Synergy-based Approach to Measuring Information Modification
Category: cs.MA
Self-Regulated Artificial Ant Colonies on Digital Image Habitats
Category: q-bio.OT
A number theoretical observation about the degeneracy of the genetic code
Category: cs.OH
Category: math.NA
Category: physics.bio-ph
Ab Initio Modeling of Ecosystems with Artificial Life
Category: physics.optics
The Enlightened Game of Life
Category: cs.DC
Artificial life, complex systems and cloud computing: a short review
Category: q-bio.NC
On Artificial Life and Emergent Computation in Physical Substrates
Category: cs.CY
That’s All, Folks!
Sorry for the sporadic linking at the end. I reached max length for a Substack post so had to stop.
That’s all for today! Thanks for reading! If you would like to learn more about networks and network analysis, please buy a copy of my book!