Author: Joe Seisdedos, SENIOR Associate & Dr Nigel Chen tan, technical assistant
The recent litigation between major Information Technology companies, such as Apple and Samsung, has reignited a number of issues around software patents. In particular, it has highlighted a noticeable flaw in the current patent system, namely the rather poor way in which software patents are classified.
The difficulty in determining whether new software infringes an existing patent or patents is well known to those within the industry. Conducting a patent search to find all patents relevant to a new concept or invention can be difficult and costly in any area of technology. However, searching for relevant software patents is especially difficult due to the cumbersome nature of the classification system used by patent offices around the world. The classification system lags behind technological development, and patent offices have struggled to revise and update their classification systems to classify software patents in a meaningful manner.
Traditional keyword searches also have a number of limitations. Certain words and phrases have become so generic in the software industry that they appear in a large percentage of software patents. It is usually necessary for searchers to spend an appreciable amount of time manually combing through large numbers of patent documents to find relevant software patents.
Most importantly, there has been an explosion not only in the number of software patents filed over the last 15-20 years, but also in the number of new products, services, concepts, ideas and solutions that have been brought to market in the software industry. The sheer volume of software available in the marketplace and the sheer volume of patents filed has created a mountain of data that is poorly classified.
However, software itself may provide the answer to the very problem created by the explosion of the software industry! “Big data” analytics has been recently proposed as a method for searching through large volumes of patent data. Patent databases contain a large volume of sophisticated data, making such databases a perfect candidate for big data analysis. Big data analytics, rather than rely on the location of a set of keywords or phrases, relies on the identification of documents that fit a certain “pattern”, the pattern being based on a statistical and semantic association.
Big data analytics is currently being developed and used effectively in a number of other roles, such as identifying customer sales trends by searching retail data, or better understanding the signs and symptoms of disease by searching patient medical records. Such systems have had a high degree of success in extracting relevant information from huge data sets. The challenge now is to use big data analytics to better understand the patent landscape, particularly in the area of software patents.
If you operate in the area of “big data” and are interested in exploring whether you could apply your knowledge and expertise to the patent world, we’d love to hear from you. Alternatively, if you work in the IT industry and you’re concerned about software patents and the impact they may have on your business, please feel free to contact us to discuss further. Our software and IT expert in the Sydney Office is Joe Seisdedos (firstname.lastname@example.org)