Skip to Content

Global Patent Map Reveals the Structure of Technological Progress

By mapping the way patents cite each other, network scientists have been able to study how different technologies rely on each other and how new technologies emerge

Navigating the web in the early 90s was never an easy business. The complaint echoed from one user to the next was that it was next to impossible to find anything.

That began to change in the mid-90s thanks to the evolution of search companies such as Yahoo. But even the early incarnations of search were difficult to use. Yahoo is a good example. Its earliest search tool was a directory that categorised webpages according to a predetermined hierarchy.

That changed rapidly when various researchers, such as Jon Kleinberg at Cornell University, realised that the network of links between websites were themselves important clues to a webpage’s significance.

The key lesson from this was that when it comes to finding stuff, the network is a far more powerful tool than the content.

So it should come as no surprise to discover that technologists are beginning to apply a similar idea to patents. Anyone who has spent a few hours navigating any patent database will know how labyrinthine and confusing they can be.

These databases are organised rather like the Yahoo directory from the 1990s– in a predefined hierarchy of topics. That certainly helps when it comes to filing patents under specific categories but it also obscures the deeper relationship between various new technologies as they emerge.

Today, Luciano Kay at the University of California Santa Barbara and a few pals reveal a new search tool that exploits the structure of links between patents to study the connection between technologies.

In their new approach, Kay and co create a network in which each patent is a node. They assign a link between two nodes if one patent cites another and define the “technological distance” between two areas of the resulting map as the strength of the links between them. So areas of this network are distant if they have few links but close if they have many links.

To test out this approach, Kay and pals have applied it to the entire corpus of patent data from 2000 to 2006 in the European Patent Office. They have also created a tool for visualising and interrogating the map that is available here. (However, at the time of writing, it did not work due to some bugs.)

The resulting patent maps provide a fascinating insight into the structure of the technological landscape and the role of various players within it.

To showcase the new approach’s abilities, Kay and co create the patent map associated with companies such as Samsung, DuPont and IBM. This shows at a glance the areas in which these companies have been working and, of course, the areas they are ignoring.

Kay and pals also show the structure of the patent landscape associated with various technologies such as graphene, nano-biosensors and so on. What this reveals is how different areas of technology are related, sometimes in unexpected ways, and how new topics are emerging.

For example, patent databases are organised in a hierarchy of families. For example patents in Section A relate to human necessities whereas patterns in Section C relate to chemistry and metallurgy. This is a somewhat artificial distinction and Kay and co say they can see strong links between various subclasses in these two sections that would otherwise be difficult to tease out.

The new map also shows the areas of densest activity. These, say Kay and co, could be powerful indicators of emerging technologies. “Patent maps may also reveal relatively unexplored technological areas that are more central to other technologies or highlight denser areas with more technological interdependency that might form platforms for the emergence of future technology applications,” they say.

That’s an interesting and powerful way to visualise a hugely complex dataset. Clearly, the existing hierarchy of technologies has its limits and a network-based approach can reveal important information about the technological landscape.

This is likely to be only a first step in a new network-focused study of patents, however. Historians of the web will be quick to point out that in the late 1990s web search underwent another revolution.

This occurred when Google introduced the pagerank algorithm to the business of web search. Pagerank assumes that a webpage is important if it is pointed to by other important webpage pages. So finding the important webpages requires a lengthy, iterative process to tease apart the structure of the network.

It’s not hard to imagine that a pagerank-type algorithm could provide some useful insights into the nature of patents and their importance. Of course, such an algorithm will need to be tuned in various ways to cope with the idiosyncrasies of the patent world.

But there’s no reason in principle why this approach would not be hugely fruitful. Indeed, similar approaches to the network of links between academic researchers and their published papers is also beginning to bear fruit.

There are other improvements in the pipeline. Kay and co say they hope to improve their mapping tool by linking the network of patents to scientific developments that are already being mapped in similar ways. That could prove interesting too.

But for the moment, those most likely to gain from this new approach will be governments and large companies who can use these tools to gain intelligence about the strengths and weaknesses of their competitors.  If that turns out to be useful, progress with this type of visualisation could be rapid.

Ref: : Patent Overlay Mapping: Visualizing Technological Distance: 

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.