LSA is Great Theory
by Roji John on July 21, 2008
Well, it may be all right in practice, but it will never work in theory. --Warren Buffett
Since its inception in the 80s (US4839853), many search applications have trended towards LSA as a method for correlating concepts. We often find that what sounds great in theory falls short in practice. In the realm of Intellectual Property (IP), Latent Symantic Analysis (LSA) seems one of these situations.
LSA can be described as a technique using statistical analysis to find associations between terms. Without getting into the mathematics, documents with similar yet uncommon terms are considered semantically close.
In theory, the methodology should be excellent for classifying text without actually reading or understanding it. In fact, for general text as may exist on the internet, LSA can prove useful in finding a handful of very related results.
Our task in the IP arena is quite different. The very specific terms used in these technical documents can often confuse LSA systems. But we also have a key advantage within IP—it’s been classified by an expert examiner at the patent office. That examiner knows precisely the concepts described in the invention.
Many users in this community are already very familiar with the US and IPC classification systems. Each system has its pros and cons. Making both classification systems available and approachable is a basic function that every patent application should provide. In addition, being able to leverage the knowledge embedded in the classification systems is key. When a document has several classifications tagged, they must all be taken into account by the application to accurately contain the invention.
Another problem with LSA is scalability. True LSA requires growing matrices that become difficult to manage and scale with large document sets. Shortcuts can be taken, but queries that are not pre-calculated can take minutes to hours to days. With today’s demand for instant information, users cannot be expected to endure such delays. Instant access to the data is a must.
Finally, LSA provides little room for user control. Its algorithms are static until updated by the provider.
I believe applications that learn from the user are better than ones that try to teach the user. Providing a responsive, accessible and repeatable method to teach the application keeps control in the hands of the expert—you!
These are all commitments we take seriously as we strive to provide the best application possible.
Passive vs. Active: What is Really Interesting about Patent Pools
by Tyron Stading on July 14, 2008
Recently we’ve seen a number of Patent Pools emerge (e.g. Alliance Security Trust Patent Pool) as a mechanism to defend against Patent Trolls. At its core, a Patent Pool is a financial vehicle designed to allow companies to pool their money and resources for either offense or defense. It won’t help you avoid litigation, make your quarterly numbers, or even release new products, but it does help to take away some of the fuel from patent trolls. However, I don’t want to talk about Patent Pools in this post, but what I think highlights a much bigger trend.
The interesting point is that Patent Pools highlight a fundamental shift I talked about in an earlier post… a shift from being PASSIVE to ACTIVE. Just 10 years ago, most patents were filed solely in hopes of one day being used to defend against litigation. Companies would sit on their patents and use the shear size of their portfolio to win battles. However, we’ve seen this is a losing strategy (i.e. RIM vs. NTP) given the current landscape. Instead of managing IP in isolation, there is a wealth of opportunities if you ACTIVELY manage your IP. With the shift toward these Patent Pools, companies are finally seeing that they need to actively manage IP around them, not just what they own. I forecast this will only increase as companies learn to take full advantage of their IP, mostly around open innovation, licensing, and partnering opportunities.
Another side effect of this trend is that it really starts to reward the inventors who are the real lifeblood of the IP system. By starting to really value patents outside the company for their impact and influence, patents are starting to truly become financial assets. Just like any other asset, they need active valuation, active management, active trading, and active utilization. Passive assets are not really assets at all and more of an insurance policy. I believe companies are starting to realize this and the trend toward active management will only increase and in new, creative ways. While Patent Pools might be a small tactic, it is just the tip of the iceberg as the mentality is shifting toward business implications.
From Patent Landscaping to Technology Intelligence
by Ryan Rozich on July 07, 2008
I get to spend a lot of time with our customers discussing their specific goals, needs and challenges related to IP intelligence. I’d like to expand on Tyron’s last post about the fundamental shifts in the IP industry and discuss what I’ve been hearing that this shift means for IP professionals and their day-to-day responsibilities.
The Evolution of IP Intelligence
I covered a lot of this in a talk that I gave at the PIUG 08 conference a couple months ago. This talk mostly described the expanding role of IP professionals and what data sources and tools can be leveraged to meet the new demands. We started off by defining some terms (both historical and new terms):
- The term patent search has been used for quite some time to describe the practice of finding patents that matched specific criteria – the business cases for these ‘searches’ usually revolved around patentability or prior-art searching.
- A while back, someone coined the term patent landscaping for looking at broader trends such as if patent activity for a given area is increasing/decreasing, etc.
- Moving beyond patent landscaping, many IP professionals are now looking to answer broader technology intelligence questions such as “what is my IP position in this market”, “who are my small/large technology competitors”, “what are the legal/market/IP threats in this area”, “who would be a good partner candidate or in/out license target”, “what are the IP/legal/other risks in partnering or acquiring this company”. These questions go deeper than simple patent searching or even patent landscaping since in order to answer those questions we need to look beyond patent data.
The Current State of Technology Intelligence
Traditionally, performing IP due diligence for M&A, assessing litigation risks, performing licensing analysis or other competitive intelligence analysis involves multiple roles all looking at the question from different perspectives. For example: a legal group might analyze the litigation landscape, IP analysts perform their patent landscaping, marketing groups look at market trends.
Each data source and group comes with its own biases and perspectives - what is needed is a common reference across all groups, competitive intelligence/marketing, legal, professional IP researchers and business stakeholders.
Building a Unified System
Having discussed the need for a unified analysis platform, we next look at what such a system might look like. Many companies have endeavored to put together internal systems to address the shortcomings of many of the vendor tools in the IP space. Most of these tools are spreadsheet-based with a lot of manual data entry involved. Here is a sample list of requirements for such a system:
- First, identify the types of data that we want to analyze and correlate and obtain the data – so the first step is to aggregate data from multiple data sources.
- The next step is that the data needs to be cleaned into a format that lets us cross-correlate the dimensions of data (for instance, we need to know that company A on this patent is the same company that we have business data for, and is the same company that we have this litigation information about), so the next step is to normalize the data in such a way that we can next correlate the many dimensions of data that we have aggregated.
- After that, we need a system that will allow us to group data, count items, and present results back in a visualized manner.
- Finally, the ultimate goal is to create a reusable system that isn’t a one-off spreadsheet for each project but instead is a unified knowledge platform that can be used across business units and roles for any project. This involves some heavy data warehousing, since data about all patents, companies, lawsuits, markets, etc need to be included.
Innography solves many of these issues and it is our goal to be a unified knowledge platform for performing these technology intelligence activities. We aggregate data from over 12 patent and non-patent data sources, normalize and correlate it and then present it in an visual, interactive format so that our users can perform advanced technology intelligence.
As the IP industry continues to mature and IP is managed like a business unit rather than legal documents, the roles of the technology intelligence professionals will continue to expand. The requirements for access to actionable intelligence will become more and more critical for companies looking to protect their IP and proactively monitor opportunities and threats.
Finding a Better Way
by Doug Miller on July 01, 2008
There is a better way, find it. --Thomas Edison
One of the things that has been most reinforced by Innography’s customers is the value of correlating multiple sources of information into a single view to gain more rapid time to results for various types of projects. Some of the use cases where this has particularly surfaced for our customers have been interesting.
Perhaps the best example is the ability to filter results of a patent query based on data that doesn’t exist in the patent records themselves. For example, for IP licensing, a technology landscape result set can be created using any number of discovery methods such as keywords, common or similar patent classifications, co-citation analysis among the most popular. That landscape can then be filtered based on company revenues of the assignee. For in-licensing a company might be interested in finding smaller companies that have some interesting innovation that might be more amenable to licensing their technology than some larger companies might. For out licensing many companies look for larger companies with a specific gap in their portfolio as a better revenue source. Both of these searches can be accomplished only if the patent landscape result set can be filtered on revenue because patent data has been correlated with company financial data.
Another case might in competitive intelligence to view a market landscape that is filtered by litigation propensity. Again a query against a specific technology can result in an interesting set of data. To gain even further insight, the ability to filter based on the litigation history of companies could indicate how protective some players in the particular market landscape are or even the market in general. In this case getting to results rapidly is even more difficult, because obtaining the litigation history and ranking companies accordingly is even more difficult, but the value of having litigation pre-correlated allows for very rapid, almost instantaneous results.
So for many that have been doing IP licensing, competitive intelligence, and patent research there is a better way, rapid results using correlated information sources. Our customers have found it. More and more prospective customers are finding it. So as Thomas Edison said--"find it.” Innography can help you do just that.
