Automation is the buzz word in any industry now. Being a patent engineer and with an experience of 10 years in intellectual property rights (IPR), I can say that automation is gradually entering in IPR domain. Some of the examples are 1) patent databases have become more intelligent in terms of semantic search and producing the results, and 2) the way patent data is visualized, etc. It is very intriguing to see where we are now and where we will go from here.
We can divide automation in IPR into three domains. 1) patent analytics, 2) patent valuation, and 3) patent drafting, prosecution, and litigation. There seem to be much progress in patent analytics, not so much in patent valuation and very less in patent drafting domain.
If we talk about patent analytics, the majority of the research is focused on deep learning and neural networks which are used for classifying/categorizing the patents and finding similar patents. Some of the automation tools are focussing on automating the report generations as well. Natural language processing is also used for suggesting contextually relevant keywords and synonyms. Some artificial intelligence tools are used for automating the patentability searches. Patent databases are utilizing machine learning algorithms to search and results more accurately.
There has been some progress on the patent valuation front as well. The reason behind not much of a success is due to patent valuation depends on both qualitative and quantitative data. Currently, the existing artificial intelligence algorithms can predict the patent value accurately for some and not so accurately for others. It is still in work progress.
Automation in patent drafting is still the initial stage because the drafting style varies from person to person and technology to technology. There is no progress at prosecution and litigation at all. As most of the lawyers and patent agents are still dependent on paperwork, even though the patent offices are encouraging online platforms
As we can infer from patent publication trends Machine learning algorithms to search and produce results more accurately that there has been an upward curve in the number of patents in IPR automation in the last five years, we can expect the same in further years as well. Coming to technology, most of the patents are focussing on machine learning models, learning techniques, and mathematical architectures.
As discussed above, the automation in the IPR domain has been growing, and it is time for organizations to adopt the technology for faster growth and explore various other opportunities in the field. It is also essential to provide opportunities to employees to learn machine learning tools and innovate new ways for patent analytics, drafting, and prosecution.
Keywords: patent analytics, automation, machine learning, drafting, litigation, prosecution, neural network, deep learning