No items found.
< Back to Blog Page

Text Comparison Tools for Assisting with Patent Claims Analysis

Text Comparison Tools for Assisting with Patent Claims Analysis

Patent analysts, conducting micro-level analysis on collections of patent documents, are taught to focus on the claims within the portfolio. The right to exclude provided by a patent covers the items found and described in the claims and, as such, it can be argued that they are the most critical part of the document. The claims can also be one of the most difficult parts of a patent to interpret for a variety of reasons, but not the least of which due to the requirements established for writing claims properly from a legal perspective in different jurisdictions.

Having a thorough understanding of what is contained in the claims and how the claims have been modified during prosecution is a key driver for being able to determine the relative value of a collection of patents and to ascertain how much coverage they provide. The ability to quickly and accurately compare the text of the claims within a patent family, on a variety of different levels, can help an analyst determine if a collection is broad and far-reaching or limited and potentially straight forward to work around

In a previous post, using an infographic to demonstrate the use of the USPTO Public PAIR system, there was some discussion on using the documents found in the Image File Wrapper tab to read the rejections provided by the examiner and to see how the applicant potentially modified the claims to overcome them. Looking particularly at a patent for a wind powered bicycle, it was shown that in order to get an allowance the applicant had to add one of the dependent claims to each of the independent ones in order to overcome the prior art. When something straight forward such as adding dependent claims to the independent ones takes place it is relatively easy to identify these changes especially when the new material is added to the end of the original independent claim. In a number of cases however, the difference in the claims can be much more difficult to identify especially when the claims begin to get longer and more complicated.

When comparing claims within a patent family an analyst can look at a variety of different levels:

  • The originally written claims (usually represented by the claims published as a pre-grant application) can be compared to the claims that were eventually allowed to grant – this analysis will demonstrated what changes needed to be made during prosecution in order to get an allowance
  • The independent claims within a single granted patent can be compared to one another – this shows the different aspects of the invention that the applicant is looking to cover
  • The independent claims from one family member can be compared to the corresponding independent claims from another member of the same patent family – through the use of Divisionals, Continuations and Continuation-In-Parts the coverage and application of the invention can be expanded

Most operating systems have built-in tools for comparing text documents. There are also a large number of third-party tools, primarily catering to the software development crowd, that allow users to quickly identify where text material has been added or subtracted from a document when compared to the original version. Another example of text comparison is the Track Changes function found in most word processing systems that allows users to identify when new text has been added, changed or simply deleted. Most users are likely to recognize examples where red lines through text represents a deletion while a colored underlining identifies that an addition as been made.

Other systems allow the user to place the text samples side-by-side and provide colors and lines to show the changes that exist between the two documents. Looking at these sorts of tools from the Apple Macintosh side of the street, Kaleidoscope is one of the most powerful and useful applications for conducting extensive text comparison analysis. At nearly $70 it is not the least expensive of the available tools, of which there are several in the Mac App Store. As mentioned, the Mac OS also has a built-in Unix based tool for comparing text files built into the developers tools but spending a little bit of money in this case provides a better experience. Using these tools to look at claim sets can be tricky so a more functional tool with a simple interface is probably worth considering. Tools for other platforms can be easily found by conducting a web search for the terms text comparison.

As an example of the different levels of claims analysis described above let’s look at some patents from a company called FitBit Inc. FitBit produces a number of activity monitoring devices including one called Flex which is worn on the wrist and monitors activity as well as sleeping patterns. In particular, there is a FitBit patent family starting with US application number 13/156,304 that we can use to illustrate different levels of claims analysis using Kaleidoscope. The ‘304 application is still pending, so we can’t use it to compare the claims from the pre-grant application to the eventual granted patent but the application has spawned several Divisionals, five of which have granted that we can use to compare pre-grant application claims to the corresponding granted patents. Looking at claim 1 of US8180591 and its corresponding pre-grant application the following is seen:

Now, while this result may seem anti-climatic, it is actually quite extraordinary, since in this case there were no changes made between publication of the claims in the pre-grant application, and what eventually granted with the ‘591 patent. This does not occur very frequently and frankly, most of the time, there are significant changes that take place from one stage to the next. If the analyst was reading the two claims without the aid of the tool it would potentially be difficult to keep them straight since there are several steps and a reasonably large number of words involved. Using Kaleidoscope the analyst is certain that no changes have occurred between the two versions.

Looking at a comparison of the independent claims between claim 1 and claim 22 of the ‘591 patent differences can start to be seen but again they are subtle and might be difficult to identify without the use of a tool.

There are significant differences in claims 1 and 22 but they are subtle. Claim 1 covers detecting motion to identify the number of stairs or flights of stairs being transversed by the user while 22 samples the motion, captures it as data and then later performs a calculation to provide the activity of the user. Claim 22 also introduces the use of a display for providing these results to the user. So while the two claims look similar clear differences between them can be quickly identified.

Finally, let’s compare the claims of the original ‘304 application that is still pending to the ‘591 patent that is a Divisional of it. This is the third level of analysis described above looking at differences within a patent family.

From this analysis we can see the potential reason for a Divisional application since claim 1 in the ‘304 application deals with the calculation of the number of calories burned as opposed to the number of steps traversed. The components of the device are mostly the same but the use of the device is different and represents a separate invention. The ‘304 application was filed prior to the ‘591 patent and a look at Public PAIR shows that it is still in active prosecution.

Reading and comparing claims between various stages of prosecution, within the same document, and within the same patent family is an essential part of understanding the scope and breadth of a patent portfolio. Using a text comparison tool, performing this task can be made much more efficient and rigorous especially when dealing with claim sets where the differences are small and might otherwise be overlooked by the analyst.

To receive our newsletter on machine learning and artificial intelligence in the field of Intellectual Property from our sister website, ML4Patents.com enter your email below:

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
patinformatics logo
linkedin logotwitter logo

©Copyright 2021 Patinformatics