Navigating the Exciting World of NLP: Staying Current with the Latest Advances

4 minute read

Published:

At present, Natural Language Processing (NLP) is having its Stable Diffusion moment, marked by rapid advances and fascinating breakthroughs. This pace of growth can be, frankly, a bit daunting. Some have even humorously suggested that Elon Musk’s idea of a 6-month moratorium on creating powerful A.I. would be a welcome respite, allowing us to better digest the wealth of new research. But since that pause isn’t likely to happen, how can we keep up with the latest in the field? Let’s explore.

Strategy 1: Leverage Specialized Platforms

Not surprisingly, there are platforms designed to help you discover new AI research papers. These sites leverage AI-powered search tools, sparing you the effort of manually browsing the internet.

Take for instance, ZetaAlpha. Their free plan offers “a powerful neural search and focused discovery across relevant open AI content on arXiv, conferences, blogs and related Github code (100 searches per month)”. It also lets you “discover trending papers on Twitter and easily find related work”. ZetaAlpha’s YouTube channel provides a treasure trove of insights on current research.

I was fortunate to collaborate briefly with Rodrigo Nogueira, a scientific advisor at Zeta Alpha, on research pertinent to the platform’s development. As part of the Neuralmind team, we made significant contributions in the field of information retrieval at the COLIEE 2021 competition. You can check out his latest talks on their YouTube channel for a deeper understanding of the research underpinning the platform.

Other noteworthy discovery platforms include ResearchRabbit and Litmap. Litmap presents an engaging graph network visualization of citations between papers, such as the citation map for the first BERT paper here.

Strategy 2: Keep an Eye on Major Conferences

Most groundbreaking research finds its way into a select group of prestigious journals and conferences. Robert Munro’s list of The Top 10 NLP Conferences is an excellent guide, not just for its list of conferences, but also for the advice on discovering relevant work published elsewhere. In his words, these are the key conferences in the field:

“To keep it simple, in Natural Language Processing you really only need to look at these six conferences:

You might also consider top conferences in related fields such as Information Retrieval, Artificial Intelligence, Machine Learning, and Data Mining:

And a few relevant journals:

Here’s a simple rule of thumb:

If a paper is published in a main conference at one of these venues or in one of these journals, it’s worth your time. If not, it might be better to skip it. The same rule applies to where you should aim to publish any groundbreaking research.”

I recommend you read the full article for more insight and possible exceptions to this rule. However, there’s no denying its effectiveness. If you follow the papers published at these conferences, and track their authors and organizations on LinkedIn and Twitter, you’re likely to stay updated on major breakthroughs just by scrolling through your social feeds.

Strategy 3: Follow the Money

While the first two strategies can help you keep up with cutting-edge research, they might not reveal progress in NLP tooling. This is where strategies 3 and 4 come into play. Most research is experimental and not immediately applicable to production - and that’s okay. However, if we’re interested in the latest tools in NLP, a more effective approach is to track the rising stars in the market.

A simple way to do this is to explore the Crunchbase page for the Natural Language Processing industry. Look for companies that have recently received seed funding, and check what people are saying about them on social media platforms like LinkedIn and Twitter. This can lead you to the latest tools being developed for NLP, whether open source or otherwise.

Strategy 4: GitHub

Lastly, don’t forget to check out trending Python repositories on GitHub. This is a surefire way to discover trending open source NLP applications. And among other resources, you can also explore the latest updated “awesome nlp” repositories.

Wrapping Up

That brings us to the end of our article. If you know of another great resource for staying current with NLP (and there are surely many we haven’t mentioned here), please share it in the comments below.