A Python Script to Uncover Creative Research Gaps Through Recombination

When I was a researcher at University, I had to come up with creative ideas at a rapid pace. Here’s a straightforward approach that may help you generate innovative concepts for your job, business, or research goals.

My particular research focus was graph partitioning, i.e., dividing a massive computer science graph structure into smaller chunks that can be distributed to multiple machines for parallel computing.

To find research ideas, I tried multiple things. One thing I tried was this simple Python script that helped me uncover interesting research gaps in my area of expertise.

💡 You can easily modify it to help you generate unique ideas in your field!

This code is writing the combination of concepts and properties to a file. The file is opened for writing, and for each concept, the properties are written with the concept appended to each line. After the last property for the concept, a new line is written to the file. This process is repeated for each concept before the file is closed.

concepts = ["GAS", "Edge-cut", "Vertex-cut", "BSP"]
properties = ["elastic", "geo-distributed", "GPU-based", "disk-based", "asynchronous"]

outfile = "combinatoricalIdeas.dat"
with open(outfile, "w") as out:
    for c in concepts:
        for p in properties:
            out.write(p + " " + c + "\n")
        out.write("\n")

Here’s the output of my idea recombination framework for enhanced creativity 😆:

elastic GAS
geo-distributed GAS
GPU-based GAS
disk-based GAS
asynchronous GAS

elastic Edge-cut
geo-distributed Edge-cut
GPU-based Edge-cut
disk-based Edge-cut
asynchronous Edge-cut

elastic Vertex-cut
geo-distributed Vertex-cut
GPU-based Vertex-cut
disk-based Vertex-cut
asynchronous Vertex-cut

elastic BSP
geo-distributed BSP
GPU-based BSP
disk-based BSP
asynchronous BSP

The code simple combines nouns with adjectives or specifiers that are relevant in my particular domain. You don’t need to understand terms such as “BSP” or “Vertex-cut” to see that these are potentially interesting combinations of ideas.

Many of them have never been researched by anybody! See the Google Scholar results of “Geo-Distributed Vertex-Cut”:

It is a 100% solid, perfect “research gap” that is open for the taking. The first paper filling it will surely contribute to the scientific community — even though it may not be super relevant.

However, if you’re not interested in rocket science, this approach to finding “research gaps” through simple recombination of concepts in your area of expertise will surely give you some interesting food for thoughts!

Here’s this same idea applied to find a great niche business:

app = ["decentralized app", "social network", "dating app"]
target = ["gay people", "toddlers", "rich people", "newly-weds"]


for a in app:
    for t in target:
        print(a, 'for', t)
    print()

The output:

decentralized app for gay people
decentralized app for toddlers
decentralized app for rich people
decentralized app for newly-weds

social network for gay people
social network for toddlers
social network for rich people
social network for newly-weds

dating app for gay people
dating app for toddlers
dating app for rich people
dating app for newly-weds

There are some really promising and fun ideas among those. Don’t underestimate the power of a simple idea recombination Python script! 🤯

💪 Action Step: Whatever your current area of expertise or interest. Create your own version of this simple yet powerful Python recombination script for inspiration!

---

Certify Your Skills