Patent Landscaping & Analysis - Call for Help

Hi Mindy,

Plan A: Check out some good Thomson Innovation webinars here http://info.thomsoninnovation.com/webinars

Plan B: Can consider checking out the next PIUG conference and to see if there is any relevant workshop that covers aspects of patent mapping (best case scenario) or at least some form of patent analysis involving thousands of patents, http://www.piug.org/2012/an12meet.php

Good day,

Rex.

Mindy,

First, I empathize with you as you have to deal with a nescient.  

Unfortunately, I do not have a simple answer for you and I know there are others in the PIUG community that have more experience than I do on this subject.  I can give you some broad ideas to consider.   Software tools can help manage/cluster large portions of data (organize and prioritize what data needs to be analyzed first), however, in the end, patents need to be read and key concepts of interest need to be categorized.  Once clean datasets have been organized, software tools (as simple as MS office products) can be used to convey trends and graphs to communicate an insight.

The first part of any good landscape analysis is to clearly define the scope, timing, and deliverables (content, format of deliverable, etc..).  If the user wants all alloys by properties ("look at all alloy patents that describe a certain set of properties and then plot out the patents that fall in certain ranges of those properties", this may not be any easy task (it more than likely will take a good deal of time to mine property data out of the patents).  Since you have 4000 results, maybe you can narrow the scope to the last 20 years and only focus on US patents.  Any project should be a balance of scope, time, and resources needed to complete the project.  Also, the whole patent landscape analysis process is an iterative process, not a one and done process and key users of the analysis should be active participants in the analysis.  Too many times an analyst will go off and do the first round of analysis, present some tables and charts, and then never hear back from the requestor again.  You need to dig for the insights.  One analysis should bring about questions for future analysis.  Knowledge management is important over the course of the analysis project.

Also, be cautious about someone using the results of analysis to answer all their business objectives.  This is not a good practice.  The results of a patent landscape analysis should not be used to answer all freedom to operate questions.

Your idea of hiring a consultant for the first project is a smart idea and one your requestor should reconsider.   As for training papers on the topic, there are many.  I have listed some below (no particular order):

1. VISUALIZE YOUR INTELLECTUAL PRUPERTY, Ronald P. Taylor and Paul Germeraad

2. Yang, et. al. Text mining and visualization tools – Impressions of
emerging capabilities, World Patent Information 30 (2008) 280--293

3. Dou, Benchmarking R&D... World Patent Information 26 (2004) 297-309

4. Mogee, et.al. The many applications of patent analysis, Journal of Information Science 28(3) (2002) p. 187-205

5. Look for articles by Anthony Trippe

Bob March

DuPont

Hello Mindy,

I agree with what both Bob and Rex wrote.  It is good advice.  In the end most automated tools will not be able to capture the level of detail you may need and as such you will have to harvest it yourself and compile it in Excel.

I am happy to speak with you and offer some additional suggestions if you want to get in touch with me.  I can be reached at 614 dash 787 dash 5237 or tony at trippe dot com (using the dash and words as hopefully a means to avoid the scapping of my phone number and email).

The theme of our Annual Conference (going on this weekend and next week) is patent analysis so while it may be too late for you to attend it is something to consider since many practitioners and vendors will be there in Denver.

I will be happy to help if I can.

Thanks,

Tony

Hi Mindy;

2 points:

1) Landscaping always involve some hard manual work - it is necessary to both understand your business needs in terms of IPR, and to grasp the scope of claim protection, not for one, but for quite a few patents. Then you have to report on the results and to place the documents on your SWOT map. You are either able to do that yourself or you are not - you have to decide.

2) different tools bring you to a different distance to your ultimate landscaping goal. Graphics may look nice, but think twice. You need to know what your bosses wants to hear - that is not necessarily just answers to their questions. Very often suggestions on how to go on strategywise mid- or long term is what they really want to hear. A tool I have came across many times and now being seriously considered in my company, because of some special shortcuts it provides in portfolio evaluation tasks, is the LexisNexis Totalpatent. Thus, without advertising, which is not allowed I think, please contact a representative and ask for a demo.

Best Regards,

Kristian

Please contact Pat Terry. We worked together on creating a analytic tool to do analysis and visualization of patents.  He's on LinkedIn.  You are welcome to connect with me and I'll introduce you.  I've been where you are so I totally understand.

Since your current project is White Space analysis, I highly recommend that you categorize your result set of 900 families (1 member per family) along one or more lines - Problem vs Solution, Problem vs matrials/alloys, Technology vs Application Areas etc., and then perform X vs Y analysis on them using Cooccurrence Matrixes (or Pivot tables in Excel). Fortunately, (and most importantly) this method can also be backed up with both logic and data if questioned by your boss. 

Since you mention, White Space Analysis, I recollect a blog article I wrote in 2008 that speaks a bit about the procedure.  Here is the link: http://blog.patentinsightpro.com/2008/10/31/white-space-analysis-using-patent-insight-pro.aspx 

The process described in the blog article can also be seen in practice in using real world technology examples in our periodic Tech Reports available at http://www.patentinsightpro.com/techreports.html . Specifically, check out the report on Fuel Additives where we highlight a sample potential whitespace.

All the best and let me know if you have any questions.

Manish

Product Head

Patent iNSIGHT Pro

Hi all,

I'm no expert, as I've used Intellixir  only a few times for large patent searches (hundreds or thousands of families). You can categorize your results using the custom fields and it automatically gives you dozens of charts slicing and dicing the data in different ways. You can also import data from several different search systems.

Monica

p.s. It does take a long time to learn how to use any of the text mining tools that I'm aware of, and even then it takes a long time to clean up the data and create the custom categorizations (weeks).

Hi Mindy,

Patent landscape has two aspects, width and depth. For a specific company or targeted sector, for example, in your company, you may focus more on depth than on width about alloy. Under such circumstance, you will need some alloy experts to help you. It is not a one-person show, but a team work. The situation of having 900 patents is not unusual. However, when a sector described by a certain set of keywords has accumulated that amount of patents, it means this direction is approaching mature. At this time, it is very likely that some of the fundamental patents (sometimes, perhaps just only one) has been secured (i.e., the biggest apple has been taken). This is about the big picture (the width) of the competition, which is far beyond the ability of any computer software can handle. Such a big picture is also an intellectual property.

If you were interested in the big picture, I can be reached by my e-mail at henry.chang212@gmail.com or phone +1-416-828-5676.

Regards

Henry

---

Muchiu (Henry) Chang, PhD. Cantab

Patent Competitive Intelligence Researcher

Monte Carlo Modeling Simulation expert

Hi Mindy,

My comment is coming very late so not sure if this will help you.

There are three critical steps to a landscape:

1. Defining the questions that you want to answer through the landscape. These questions will determine how broad you need to be with your search strings and how deep you need to be in your analyses. White space itself is a very broad area, and the way to narrow would be to understand the business objective. You can do a top level analysis based on IPC, ECLA or US classes, and most commercial databases provide tools to do these basic analysis. If you need to do a deep dive, then there is no escaping from manually analyzing and categorizing the patents - however, defining the questions can help you in reducing the set to a manageable scale.

2. Defining a taxonomy or tags for categorizing or tagging patents - This has to be in collaboration with the end user, and is constantly evolving during the course of the analysis. You can start with a basic taxonomy - get the end user input, and then refine it as you analyze more patents and start identifying technical trends. It is important to understand whether you want to restrict the analysis to claims or expand it to full-text. You will need to create a database to store this information is a systematic manner - I find excel to be sufficient for most cases.

3. Analysis and visualization: This is where you present your insights to the end-user. Many of us forget that the purpose is to convey insights and not create fancy charts. Sometimes, just putting the information in a table is sufficient.

What is really helpful is if you can build a mechanism to search the "database of relevant patents" that you have tagged/categorized. This will allow you to run various kinds of scenarios - do analysis by specific sub-domain, by specific competitor, etc for more meaningful insights. If you can get help from your IT team and develop a search macro for the excel database that should be sufficient.

If you have any questions feel free to call or email me at 610 dash 458 dash 2179 or ashish dot nawani @ evalueserve.com

Regards,

Ashish

Note: These are my personal comments and are not endorsed by my employer

Hi Mindy,

Another great patent landscaping alternative in the market, the Business Intelligence module from Orbit.com:

I kept meaning to respond sooner and time just slipped away from me! I knew I would get great advice from this group, but I am admittedly blown away by the overwhelming response – thank you all so much! Between the personal e-mails and comments there was a lot to digest, but I have a better grasp on the situation. I have made some recommendations to management on where we need to take this to get to the next level with landscaping. It means the world to me that I could turn to a group of experts and get quality, honest advice. I am very grateful for all the time and effort you all put in to your comments and suggestions.

Thanks for everything,

Mindy Peters

Hi Mindy,

I'm glad you found the PIUG community helpful. One point I wanted to add is that many of the challenges you are facing are the same challenges faced by experts in the field. Being able to translate a request for patent analytics into a good deliverable can be difficult when the requester is not an IP expert. Every requester will have a different objective for their deliverable, and sometimes what is requested does not match up with what they really want or with what is possible for a given budget and deadline. Even the most experienced IP professional will need to ask many questions and manage expectations in these cases. Good luck!

Anders