Thread by @danofer

1/ 🧪🔬 Ever wondered if scientific research can be predicted? Our latest paper "What's next? Forecasting scientific research trends" tackles this question. Check out the details here 👉 arXiv:2305.04133v1 [thread] 🧵

2/ 🌍 The world of science is constantly evolving. But what if we could predict these trends, especially in life sciences? That's what our work aims to do. Imagine the possibilities!

3/ 📚 We've mined historical publications, research/review articles, and patents. The result? Models that can foresee scientific trends 5 years in advance. Time travel, anyone? #AI #DataScience #timeseries #trends

4/ 👀 We found an interesting pattern - with a lot of review articles are often going to start declining.
A great insight into how scientific topics ebb and flow over time. 🌊

5/ 🤖 Language models lent us a hand in this journey, helping us understand topic temporal dynamics across molecular, technological, and conceptual areas. A fine example of AI aiding scientific exploration! #NLP #AI
Thanks to #sentencetransformers @Nils_Reimers !

6/ 💡 Ever heard of CRISPR-CAS9 gene editing tech? Patents on this tech predicted future research papers. Patents as a crystal ball - who would've thought! #CRISPR #patents #patenttech #uspto #patentsview

7/ 📊 Our work isn't just an academic exercise. It has real-world implications for early stage researchers, educators, and policymakers. Knowing what's on the horizon in science can guide career development and resource allocation.🌟

8/ 🔮In conclusion, our work shows that with the right tools and data, we can peek into the future of scientific research. It's not perfect, but it's a start.#FutureOfScience

9/ 📝 Read the full paper for all the details and let us know your thoughts! 👉
arxiv.org/abs/2305.04133
"Whats next? Forecasting scientific research trends"
@danofer and @LinialMichal
#ScienceTwitter #ResearchTrends

Also, a big thanks shoutout to @edsperr for his website esperr.github.io/pubmed-by-year/ - it saved us a ton of work in cleaning the data!
#pubmed #pubmedbyyear

Code and full pipeline/models are free here: github.com/ddofer/Trends #opencode #github