Reimagining IPR: Relecura’s AI-Powered Journey
By Mamata Saha
While we’ve often heard the saying, “Change is the only constant,” we’ve never actually mulled it over despite sustaining in a world where the statement is corroborated every day. Likewise, the realm of Intellectual Property Rights (IPR), is no different. With innovation and technology progressing at breakneck speed, the field of IPR is going through a paradigm shift and redefining ways of creating, using, and protecting intellectual assets in a world identified and connected by technology. Finally, data and algorithms, with substantial contributions by technologies such as Artificial Intelligence and Machine Learning, have paved the way for unique and unmatched insights and functionalities.
Relecura’s AI-powered tools: Reshaping Intellectual Property
Relecura, an AI leader in the field of patent search and analysis, has, since its inception, maximized the potential of AI and ML, to build tools for the resolution of IP-related requirements and issues, patents in particular. Among the tools, some worth mentioning include Novelty, Taxonomy Builder, and the AI Classifier. Let’s take a few minutes to go over each of these.
Manual classification of huge datasets is a drain on time, resources, and money. However, Relecura’s automated and AI-powered Classifier does the job in far less time with precision and accuracy levels that could never be achieved manually. The tool uses the input of a few training documents to create AI classification models, which are then analyzed and the documents are classified into specific buckets. Essentially, it uses a three-step process consisting of model creation, model analysis, and document classification to categorize large numbers of patent documents. Last but not least, the tool’s confidence metrics share key data on how well a document matches a particular class.
Let’s consider a scenario where an FMCG company has a massive patent portfolio that it needs to sort, organize, and classify. It decides to use the Classifier to get the job done. As shared earlier, first, a model has to be created using ML algorithms, for which the tool needs to be trained using a few documents from the portfolio. In the image below (Fig 1), you can see that once the documents have been shared (either by importing a file or through a search query on the Relecura platform) and trained, the tool retrieves the key categories (i.e., Detergents, Soap, and Hand Wash) of products that form part of the patent repository, along with the number of positive and negative documents. Positive documents are those that belong to the relevant category while negative ones do not. Here, we see only positive documents for all three categories.
Fig 1: This image depicts the model creation process of the AI Classifier.
Once the model is created and studied, you get to see a branched out tree-like taxonomy of the classification of the patent documents. You see four categories, Detergents, Hand Wash, Soap, and Others, and the patent document count for each. For Detergents, the count is 378, for Hand Wash, the count is 20, and for Soap, the count is 15. The ‘Others’ category comprises those documents that don’t belong to any of the three categories.
Fig 2: The branched out tree-like taxonomy, classifying the patent portfolio into its respective categories.
We come across many examples of brilliant ideas every day, but not all of them may be unique and patentable. To gauge whether an idea is non-obvious and commercially viable, we need an efficient tool like Novelty. Novelty is an invention analysis tool used to examine the uniqueness and patentability of an idea or invention. While the only input needed is a few paragraphs highlighting the invention, AI and ML algorithms work in the background, analyzing the text and summarizing it graphically through keyword connections. Finally, the tool delves into a global patent database of over 160 million documents spread over 160 geographies to fetch the relevant data. Below are some images to describe how the tool functions.
Fig 3: The user interface of Novelty, where a user can input a few pieces of text and click on Submit. In this image, a few lines of text describing the creation of a smartwatch, have been inserted into the box.
Additionally, with the help of a patent number (if the user has access to one), they can run a prior art search or even access the abstract or first claim of a specific patent if they’re interested in the information.
Fig 4: This image depicts the graphical representation of the smartwatch invention highlighted in figure 1.
Using keywords from the text shared by the user, the tool was able to put together a summary graph of the invention, which in this case is the smartwatch.
When patent portfolios contain a large number of sub-technologies, managing the data is difficult and time-taking. However, with Taxonomy Builder, data asset management becomes much simpler and quicker. Taxonomy Builder is an advanced, automated, AI-driven tool that helps organize, categorize, and manage large patent portfolios. The categorization results in a branched-out taxonomy tree, depicting the technologies and sub-technologies in an easy-to-read and aesthetically pleasing manner. Interestingly, the tool can also help you determine the white spaces in your technology and patent portfolios, thereby helping you figure out your requirements for additional patent or technology acquisition, identifying patents for licensing or selling, and even fetching potential leads for it.
Fig 5: This image depicts the entire patent portfolio of Apple, the American IT giant, comprising 167,815 patent documents. It shows the multiple sub-technologies under Wireless Communication Networks and their respective document counts.
In continuation with the above image, here’s another one showing another technology, Digital Computing Systems, its many sub-technologies, and the document count for each.
Fig 6: This is the second part of the taxonomy displayed in figure 5 above, where the sub-technologies under Digital Computing Systems and their respective document counts are depicted.
Courtesy of the digital revolution, IPR is evolving and will continue to do so in the days to come. Safeguarding and harnessing the power of IPR will depend on the tools we use, the action plans we implement, and the insights we gather along the way. With its tools driven by AI and ML, Relecura has already become an integral part of the digital transformation process. In fact, with the help of its predictive analytics, data-driven decision-making, and automation, Relecura is playing a pivotal role in giving form to what the future of IPR will look like!
