AI Versus Manual Patent Searching: How a Hybrid Approach Can Optimize Success

By Sumit Prasad
October 2, 2021

“Although AI-based patent search tools are agile and user-friendly, they can be integrated with manual searches conducted by experienced analysts to obtain better results.”

AI - https://depositphotos.com/144709683/stock-photo-artificial-intelligence-concept.htmlWith the forecasted growth of global Artificial Intelligence (AI) market size, it is evident that AI is quickly becoming the solution to most software and service needs. AI has even infiltrated our homes—for example, we are increasingly seeing smart home systems that incorporate Internet of Things (IoT) technology along with a master AI virtual assistant.

Undoubtedly, the technology has made space in the intellectual property-based service sector as well. For instance, to support patent searching, there are quite a few AI-based automated patent search tools available. Although many of these are still in their training stage, these tools are likely to mature. Thereafter, the question looming over innovators is whether to take advantage of affordable AI patent search tools or invest in outsourced manual patentability searches.

With AI-based patent search tools being available on subscription-based models at affordable prices, the dilemma mentioned above becomes even more imperative. So, although AI-based patent search tools are agile and user-friendly, they can be integrated with manual searches conducted by experienced analysts to obtain better results. The following article draws a comparison between both options and defines scenarios where these models can be deployed.

Manual Patent Searching Still Wins Out Overall

  1. Exhaustive Searches – Artificial intelligence and machine learning (ML) algorithms are still in their rudimentary stages of training and development. Therefore, they are still unable to perfectly replicate the full extent of abilities of a real human analyst while performing exhaustive searches. Therefore, human intelligence can be deployed to optimize results.
  2. Opinion – Although AI-based patent search tools can give innovators a quick idea of whether to proceed with an innovation, they do not provide a complete analysis of all the white spaces of innovation in the particular patent landscape. Manual searches can provide innovators with such an opinion.
  3. Improved Quality – Moreover, to improve the quality of claims based on prior art, innovators seek patentability opinions for their drafted claims. Therefore, manual searches are also required to improve the quality of the patent application.
  4. Reduced Time – From invention disclosure forms (IDF) to patent filing, manual searches are less time-consuming for independent claim drafts. However, with respect to speed generally, automated searches are always faster.
  5. Seamless Refinement – When it comes to refinement of search strategies based on the particular subject matter, human IP analysts with expertise in the respective discipline have an upper hand. The human element often adds refinement to patent search results. However, in automated searches, we typically just get a list of outcomes and no relevant analysis to accompany it.
  6. Native Language Searches – Since AI has not yet been trained in several native languages, patent searches in this case are only possible with manual intervention. Translators and native language speakers need to be involved in such patent searches.
  7. Identifying Non-Patent Literature – Automated searches are still not competent enough to search for non-patent literature in different disciplines of technology. Therefore, in this context, manual patent searches are much more comprehensive and accurate.
  8. Learning from Prosecution History – In a scenario where invalidity searches are needed, we need to learn from prosecution history of the nearest prior art of the technology being developed. Analysis of these—which must be performed by a human—can help in identifying unique areas of innovation that can be used further for innovative product development.

When to Choose Automated Search

  • Quick Validation: Automated searches can be used to quickly validate concepts at the ideation stage. Such AI-based automated search tools give innovators access to existing prior art so that they can build their innovations around such existing patents.
  • Prior Art for IDS Submission: When submitting IDS, relevant prior art needs to be submitted to validate the innovation. Automated searches can be used to include prior art in IDS and to reduce the turnaround time for patent prosecution.
  • New Technological Domains: Inventions that are unique and revolutionary often pose an issue to the in-house subject matter experts (SMEs) during the search for relevant new prior art. In such cases, the subject may not have any direct prior art, but AI-based patent search tools can be used to understand and derive similarity from the latest technologies in order to identify potential existing prior art.
  • Cost Reduction: Since most AI-based patent search tools are relatively affordable, there is no requirement to maintain a budget for outsourcing patent search services. Many AI-based patent search tools use a pay-per-use model for patent searches.

When to Choose Manual Search

  • Crowded Patent Spaces: For patent spaces that are crowded, manual searches can help in differentiating between the existing patents and prior art for the given invention. Since the subtle differences in such prior art give rise to patentability, it is rather beneficial to differentiate such subject matter with respect to the available prior art.
  • Interchangeable Terminologies: Similarly, such crowded patent spaces often have terminologies that are overlapping with those of the invention. Some technology domains have terms that are used interchangeably. To perform an exhaustive prior art search, all such terms need to be searched manually – in a way that is relevant to the given context.
  • New Application Areas: Inventions that are not novel by themselves, but are novel in their application, require extensive manual prior art searches to identify any prior art similar to the given application. By doing this, claims can be drafted in such a way that ensures high licensing value in the future.
  • Contemporary Searches: With a continuously expanding patent market and increasing competition in patent filings, the need to understand state-of-the-art (SOA) in terms of validity has become quite important. Therefore, contemporary searches that incorporate both a novelty and an SOA search are necessary to understand competitors’ patent filings. This also allows innovators to draft their patent claims by adapting their patent filing strategy with that of competitors.
  • Avoiding 103 Rejections: To avoid obviousness rejections under 35 U.S.C. §103, manual searches can help innovators to analyze the patent prosecution history of similar patents. This can help them to learn from the prosecution strategies of patents that have similar combinations of technologies. Such a customized patent search can thus ensure that the chances of getting a 103 rejection are reduced.

Complementary Use of Automated and Manual Searches – the Hybrid Model

For organizations with large research and development (R&D) teams working on numerous innovations simultaneously, manual searches performed at the initial ideation and concept approval stages can be time-consuming. AI-based tools solve this incumbent problem by providing quick prior art search results so that such R&D teams can leverage the role of IP in their processes. Even though most of these AI-based tools are still in their nascent stages of training and input data, they can complement manual patent searches in certain cases. While AI-based tools are economical and quick, manual searching is more reliable and relevant. Using a combination of both can ensure truly superior results.

 

The Author

Sumit Prasad

Sumit Prasad is an Innovation Strategist by profession with more than 8 years experience as a patent expert and IP consultant. He is playing an instrumental role in cultivating innovation and spreading IP awareness among Startups, MSMEs, R&D groups of GCoEs and helping them in strategizing their IP activities. At Sagacious, he is leading the ‘IT for IP Initiative’ that allows him to leverage his IP experience to create algorithms using AI, ML, NLP, Automation, etc. and develop tools/software that can help solve problems in intellectual property industry. Sumit has also helped Sagacious in automating IP processes and development tools for enhanced efficiency and better interaction with clients.

Warning & Disclaimer: The pages, articles and comments on IPWatchdog.com do not constitute legal advice, nor do they create any attorney-client relationship. The articles published express the personal opinion and views of the author as of the time of publication and should not be attributed to the author’s employer, clients or the sponsors of IPWatchdog.com. Read more.

Discuss this

There are currently 3 Comments comments. Join the discussion.

  1. Sankar Sundaram October 4, 2021 12:31 am

    Also, there is an opinion that, for the Primary Classification of a patent, a human being is more intelligent than a machine – at least as of now.

  2. Yojit Bhugra October 11, 2021 6:50 am

    This was helpful in explaining the AI-based tools and the limitations that we have with these tools. Nice Work

  3. Jacky November 4, 2021 10:47 am

    Excellent objective analysis.

Post a Comment

Respectfully add to the discussion.

Name *
Email *
Website