Patent mining is a process of searching and analysing patents for further extraction of insights related to innovations in a specific domain by using various patent databases. These analysis can be helpful for companies and organizations, to gain insights into the technology trends, competitive landscapes, and potential risks. The information that results from patent mining can be used for various purposes, such as innovation, strategy, licensing, and/or litigation.

Patent mining involves a set of activities including analysis of patent publications, reviewing different sections of patent publications, evaluating patents based on a scoring system, and investigating patent-related papers for a specific purpose.

Innovative companies that have been in business for a long time are likely to have a large number of unaccounted IPs, which they may be unaware of. These IPs may be converted to performing assets, if utilised optimally and timely. However, large organisations may often find such an exercise cumbersome and may end up losing out or abandoning important IPs in the public domain. Such suboptimal IP asset management practice may affect the overall business of the company. Therefore, specific analysis and studies are regularly conducted by such companies to classify the IP into various categories to indicate its important to their business goals. Such an exercise is an integral part of due diligence analysis  during a merger or an acquisition of companies.

A typical mining exercise involves firing keywords/key phrases in databases to discover relevant patent documents. Next, the patent documents are strategically analysed and systematically segregated/classified for ranking based on a set of predefined parameters. Such parameters may correspond to technical features, application, business significance, claim scope and jurisdiction, prosecution history complexity and the like. The classified patent documents and data mined from them may be compiled and presented in the form of graphs, slides, technical paper, excel sheet etc. The search and analysis exercise for Patent Mining is conducted by a dedicated team of professional in a particular/relevant field.

Data Collection, Data Cleansing, Data Analysis, Data Visualization, and Data Reporting on the Patent.

The individual steps involved here delve into as

Data Collection

The relevant patent documents are searched for in various patent websites/databases using searching tools to gather the information relating to the domain.

Data Cleaning

The unwanted data(noise) is removed during the data cleansing step, where, according to the needs of the clients, the data is filtered based on a set of predefined parameters.

Data Analysis

The curated data is further analyzed and classified based on a ranking criteria. The ranking criteria may include factors such as patentability, the litigation involved, possible infringing, IP valuation, monetizability, and other relevant factors using a a mix of statistical and computational processes.

Data Visualization

The analysed data is then visually represented in a graphical format and inferences are added wherever applicable.

Data Reporting

The final step involves reporting in a format best suitable for client’s request.

Various tools and techniques are employed for effective patent mining. These tools include patent mining software such as VantagePoint and Thomson Innovation that are adept at analysing big datasets and implement Machine Learning Algorithms that can predict trends and classify patents based on various input criteria. Natural Language Processing (NPL) Techniques are also implemented for extraction of information relevant to the patent texts and identification of key themes. Further, visualization tools like Tableau and Power BI are utilized to generate the visual representations for creative interactions and insightful visualization.