While seeking patent protection for any invention, it is always advised to perform exhaustive patentability searches. This is because this search is the first crucial step for innovators or patent attorneys to evaluate whether the respective invention fulfils the patentability criteria or not. Furthermore, patentability search is a time-consuming process as it involves going through hundreds of patent and non-patent literature to identify prior-arts that might affect the patentability of the invention.

Resultantly, decision-makers often have a difficult time deciding whether to opt for an artificial intelligence (AI)-based patentability search or a manual one, depending on their objective. The following article unveils the pros and cons of both manual and AI tool-based patentability search and suggests the best strategy that can be relied on to minimize any setbacks during patent prosecution or examination stage. So, let’s understand the problem first.

Understanding the Dilemma

Innovators around the world filed 3.3 million patent applications in 2018, up by 5.2% for the ninth consecutive yearly increase. As a result, high patent application count and annual growth in number of patents has necessitated more time and resources for executing patentability search efficiently. Thus, to speed up the patentability search process, the world has started looking toward AI tools i.e., artificial intelligence-based patent searching mechanisms that promise to reduce time investment in searches from 2-3 days to a few minutes. With a speedy outcome of the patentability search, stakeholders can promptly plan their next steps in terms of moving forward with patent protection and launch of the product in market. However, this poses a big question – “can decision-makers solely rely on existing AI tools to save time?” The points below will help in answering this question.

Manual Search – Pros and cons

Advantage – Human Intelligence and Expertise : In manual patentability search, the entire strategy is prepared by a patent searcher. He identifies the keywords, classes and prepares the search strategy. Thereafter, the searcher screens the captured results to identify promising ones that might act as prior-arts and challenge the patentability of the invention. This is an effective and efficient search process as it relies on the intelligence and experience of the patent searcher. Also, manual searches minimize the chances of missing out on critical prior-arts.

Disadvantage – High Time and Resource Investment : Manual patentability searches consume significant time and resources because, quite often, sizable data needs to be screened. For fast-paced inventions where the shelf life is short, for example, trendy toys or fashion accessories, these searches might act as one of the factors that can cause a delay in market entry.

AI-Tool-Based Patentability Search – pros and cons

Advantage – Highly Automated Process and Faster Results: AI tools rely on deep learning, neural network, and natural language processing to create patent search strategies and analyze the huge volume of information automatically. One just needs to enter the simple description in the AI tool interface to initiate the patentability search. The rest of the tasks like preparing the search strategy and shortlisting of the top results is carried out by the AI tool itself. All this is performed in merely two-three minutes.

Further, the AI tools learn from the prior process data to improve their overall search accuracy. With AI-based patentability search, one can identify the prior-arts with reduced speed and time.

Disadvantage – Lesser Feasibility: A study has suggested that none of the available AI tools can support every aspect of the prior-art search process. The experimental results for precision varied between 30% and 50%, which means that the first 10 search results contained between 3 and 5 relevant documents.

AI Tools and Human Intervention

There are AI tools in the market that allow human intervention in order to improve precision of the search process. One can feed the important aspect of the invention, choose the keywords or classes to be used, set the count of top results to be displayed, etc. in the AI tools. Sagacious IP also keeps an eye on such AI tools and tests the promising ones to find out their real-time efficiency in order to serve organizations better. An efficiency study on one of the shortlisted AI tools, conducted internally at Sagacious IP, reveals that AI tools can perform better with human intervention rather on their own.

For simple consumer-related inventions, AI tools can capture the relevant prior arts with minimum human intervention, which was limited to selecting the key concepts only. However, for complex industrial inventions such as machine internal structure, AI tools were able to capture few related results that were close to the subject invention. The tools could do so with dedicated human intervention that included selecting the key concepts, adding the synonyms, manually feeding one or two domain-related patents, etc.

Obtaining the Best of Both Worlds

To conduct an effective patentability search, integration of both patentability approaches is suggested. Both manual and AI-based searches can be performed sequentially in two stages. The initial search and screening can be conducted using AI search tools and in case of an unfavorable outcome, one can complement it by going forward with the manual patentability search. This integrated patentability approach can use accuracy and effectiveness of manual patentability search and quick turnaround time of the AI-based searches to ensure identification of quality results at considerably lower costs and turnaround time.

Sagacious IP patentability searches help businesses to verify the novelty of inventions and draw optimal patentable subject matter. We use highly effective AI tools in combination with manual searches to offer best results to organizations.

• Rohit Kumar, Nitin Sharma (Engineering Searching) and the Editorial Team

Having Queries? Contact Us Now!