My password to the private beta for WestlawNext is set to expire tonight, but before it does, I have some final thoughts on the platform for you to consider.
First, the browser. If you are running IE6, you will need to upgrade to IE8. If you are running IE7, you’ll get by, but there are definitely a few glitches I’ve experienced, so you’ll probably want to upgrade. Best performance currently is with Firefox, but IE8 runs a close second. Last week we were told that Google Chrome is the most stable browser for WestlawNext, but that won’t be rolled out until February 18th. For most of you, however, that won’t matter because you won’t see this thing until much later anyway. But why is this important? It’s just another cost to adopting the platform. If you need to upgrade, then you just have to factor in the IT time to do so. I’m still surprised at the number of firms running ancient-ass browsers. Really folks? Do you not care about online productivity?
Second, secondary sources. I started thinking about the fact that WestlawNext gives you access to over 6,000 secondary sources. And while that is a big number, it’s nothing compared to the primary source material made available, which is to say, search for secondary sources probably could be a lot better if it was a stand alone product. I suspect that relevancy in secondary sources could be much better if all a user was doing was searching those databases with WestSearch. I’m actually thinking there should be a two-step portal, where WestlawNext tries to help you find an explanation to your question first before sending you on to discover the answer for yourself through primary material. This first portal could also be optimized for reading secondary sources that is different than how we try to consume primary data.
Third, WestlawNext SEO. I didn’t think of this, I think Greg Lambert did. But after playing with the system, there are some inevitable questions as to whether the WestSearch algorithm could be gamed by users. If WestSearch is updated weekly with customer log information, as was indicated by the design team, then there must be some sort of potential for increasing the rankings of certain cases over others. Why would this be important? Well, I think Greg posited that if you found a case more favorable to you on a point of law, you might be able to increase its relevancy (for other users looking for the same or similar information) by creating artificial “meaningful relationships” (see my last post on this topic), such as, placing the case in your folders, printing, etc. You could actually encourage the entire firm to do it as a matter of practice. I’m not sure it would work, but it’s certainly something worth considering as more consumer-user data is incorporated into the search algorithm. Lastly, there’s also the problem of law students to consider. Once this platform is rolled out academically, I think it is fair to ask whether law students’ log data will be added to the algorithm mix. Honestly, I think they could screw it up faster than the concerted effort of a few lawyers.
Fourth, social tools. I’ve stared at this platform, and while I know that WestlawNext will be able to let users “vote” and “make comments,” that functionality appears to be specifically limited to the user communicating with Thomson Reuters Legal (TRL) through the “Next” link off the home page. It’s basically a way for you to tell TRL what you’d like to see in later releases. But that is it. An entire community of lawyers across the country, or in the same state, or same city, logged in at the same time, cannot communicate with each other, at all, other than through shared folders and sticky notes. I’m not belittling those advancements, which are important. I’m just puzzled why a company that wants to increase engagement for their site doesn’t find a way to leverage something no one else has, namely, the brain power of tens of thousands concurrent users online at any given moment. If you could connect to someone or more than one person reading the same case you were, would you? Would you ask, What does this passage mean to you? What if your identity was kept anonymous, or just showed Jill, Houston, Texas? Could we—can we—have meaningful conversations about the law in real time, or is it a waste of that precious resource? What has Twitter, or Google Wave for that matter, taught us about our capabilities for social?
Finally, search routine. Like most legal researchers, I’m a Boolean user. So, it is hard for me to abandon it. But I was committed to giving WestSearch the benefit of the doubt, so I always defaulted to the standard “plain language search.” Most of the time I was given relevant results, and my initial research expectations were met. However, I wasn’t satisfied I had gotten everything. So I would scan the first 10 to 15 entries, perhaps viewing 8 of those, and then narrow my results using Boolean in “Search within results,” probably two or three times with different queries. Only then would I be satisfied with what I’d found. This is important because it changes the value equation. I suspect that most users will be like me and feel compelled to narrow the results using Boolean. Now, I know that filtering (e.g., court, published or unpublished opinion, date range) is not an extra QuickView charge, but I don’t know about “Search within results.” You should probably ask your rep because if it does cost extra, then the value proposition for the service is also changed.
[Image (cc) by Kaptain Kobold]
At the beginning of this week, I was fortunate enough to get to spend some quality time with fellow legal information professionals at the Thomson Reuters headquarters in Eagan, Minnesota to discuss the history and future of the new Westlaw, which we now all know as WestlawNext (on February 8th, it will be found at “next.westlaw.com”). [Disclosure: Thomson Reuters paid for the flight, hotel accommodations, and incidentals.] And over the last two weeks, I have been lucky enough to use WestlawNext in my own research, and have learned a fair bit about its capabilities beyond what we were told this week. If you haven’t seen Jason Eiseman’s video of me, the always entertaining Tom Boone, and the desperately in need of hair coloring Greg Lambert, at Jason’s blog or on Henderson Valley Eggs, I suggest you skip over there for a bit and spend some time watching it because the video is perhaps the most honest, off-the-cuff assessment of the product you’ll find on the Tubes right now.
QUICK TAKE.
I like it, a lot. Thomson Reuters Legal (TRL) has put together a pretty solid offering that tackles, in my estimation, what Seth Godin calls the “paradox of the middle of the market:”
The middle of the market is a paradox because of the inherent contradiction between the ease of reaching the nerds and the geeks and the need to reach the middle. The solution, if there is one, is to enter a market to the enthusiastic cheers of those in search of the new, but to build a product/service that appeals to those in the middle. After the initial wave of enthusiasm, you hunker down and ignore those that first embraced you, obsessing instead on the needs and networks of the middle. It’s a difficult balancing act, but it’s the only one that works. Seth Godin, The paradox of the middle of the market (June 28, 2009)
WestlawNext does a good job of satisfying those who want a new, modern interface with some Web 2.0 bells and whistles without sacrificing core functionality. Not everyone is going to be satisfied though, and that’s the point because most users will. And the loudest voices won’t come from the nerds and geeks wondering where all their social media tools are, it will come from the established among us, the Boolean diehards. In fact, if you listen carefully now, you can already hear their rising cacophony of shouts, “Give me Terms and Connectors, or give me Death!”
Forgive me if I’m late to the party on this one as the debate seems to have been going on for the last decade (at least in some circles). Last week Physorg posted about an article in the inaugural issue of WIREs Cognitive Science, titled “Neurolaw.” The following passage caught my attention:
One of the most controversial ways neuroscience is being used in the courtroom is through ‘mind reading’ and the detection of mental states. While only courts in New Mexico currently permit traditional lie detector, or polygraph, tests there are a number of companies claiming to have used neuroscience methods to detect lies. Some of these methods involve electroencephalography (EEG), whereby brain activity is measured through small electrodes placed on the scalp. This widely accepted method of measuring brain electrical potentials has already been used in two forensic techniques which have appeared in US courtrooms: brain fingerprinting and brain electrical oscillations signature (BEOS). Brain fingerprinting purportedly tests for ‘guilty knowledge,’ or memory of a kind that only a guilty person could have. Other forms of guilt detection, using functional magnetic resonance imaging (fMRI), are based on the assumption that lying and truth-telling are associated with distinctive activity in different areas of the brain. These and other potential forms of ‘mind reading’ are still in development but may have far-reaching implications for court cases.
A quick search of Westlaw unearthed a number of cases on “brain fingerprinting,” and for those courts squarely faced with the issue, the results appear to be similar:
Brain fingerprinting purports to measure certain patterns of brain activity to determine whether the person recognizes or does not recognize offered information, thus revealing what the person has stored in his or her brain. We agree with [the trial judge] that no authority exists in this jurisdiction to establish the reliability of such testing, and the results of any such testing would not be admissible in evidence. And, defendant did not proffer any competent expert opinion to establish the reliability of brain fingerprinting evidence. (Citations omitted.)
State v. Bates, No. 34-2007 (Sup. Ct. N.J. 2009) (unpub.; 3-23-09). Many of them cite to Harrington v. State, 659 N.W.2d 509, 516 n.6 (Iowa 2003), which characterized the process as “novel computer-based brain testing:”
This testing evidence was introduced through the testimony of Dr. Lawrence Farwell, who specializes in cognitive psychophysiology. Dr. Farwell measures certain patterns of brain activity (the P300 wave) to determine whether the person being tested recognizes or does not recognize offered information. This analysis basically “provide[s] information about what the person has stored in his brain.” According to Dr. Farwell, his testing of [the defendant] established that [the defendant]’s brain did not contain information about [the victim]’s murder. On the other hand, Dr. Farwell testified, testing did confirm that [the defendant]’s brain contained information consistent with his alibi.
It does not surprise me that I have been unaware of neuroscience developments as they related to guilt or innocence in the law, as I spend almost all of it in the civil arena. But I did find several helpful articles on the subject, one being a note from the Yale Journal of Law & Technology, Roberts, Everything New Is Old Again: Brain Fingerprinting & Evidentiary Analogy, 9 Yale J. L. & Tech. 234 (2007), another by Deborah W. Denno titled Crime & Consequences: Science & Involuntary Acts, 87 Minn. L. Rev. 269 (2002), and an article from the NYT by Jeffrey Rosen titled The Brain on the Stand. Unfortunately, in the time I alloted myself, I couldn’t find anything on fMRIs or other forms of “guilt detection.” This seems like an interesting area of the law, and I’ll be curious to see if it progresses beyond junk.
[Image (cc) laimagendelmundo]
Through a glass door you watched
the surf break in the distance
while someone told you confidentially
that what’s really important
is the way time passes in dreams
when you’re not dreaming.
Desperate Characters, Nicholas Christopher
[[Photo (cc) bortescristian]]
We are on the eve of Apple’s “Come See Our Latest Creation” event, and the only thing I am thinking about is whether the fabled New Device will be a significant advancement beyond my Old Device, thus rendering all previous devices completely undesirable and highly repellent. And from what I’ve read by Tim van Damme, John Gruber here and here, Marco Arment, John Siracusa, Aaron Mahnke, Andy Ihnatko, Neven Mrgan, and Jim Darlymple, it will be an amazing Latest New Device. Yet, I’m still unsatisfied because none of them talk about how the Tablet will usher in a new vision of how we will read. The closest is John Siracusa, who observes,
[The Apple Tablet] will provide an easy way for people to find, purchase, and consume all kinds of media and applications right from the device. It’s that simple. … Apple’s doing the hard work to make all of this happen, of course. That means courting a new class of content owners whose wares are a good fit for a tablet-scale device: print publishers. Apple’s got a lot to offer publishers: millions of existing customers who’ve proven their willingness to buy digital media, relationships with other big media companies to show that Apple knows how to get along in this world, even a CEO who is himself the head of a movie studio and the largest single shareholder of a media giant. Add to that the color, video-capable touchscreen, which current electronic publishing suitors lack, and Apple can now appeal to new kinds of publishers: glossy magazines, comic books, and mixed media hybrids (e.g., People magazine with embedded celebrity videos).
John Gruber over at Daring Fireball has another excellent post, this time on the Apple Newton. In considering why the Newton failed, Gruber says this:
I think it’s OK for a 1.0 product to be ahead of its time, to be too ambitious. The trick is not to be too far ahead, and more importantly, for the follow-up products to improve practically.
When it comes to products generally, I wholeheartedly agree with him on this point. If you’re too far ahead, the consumer will never catch up. But when I survey what is happening in legal publishing, I have what Dan Visel says is “the sense of being on the verge of a future that’s dragging its feet.” Our industry (like science, according to Michael Clarke) has not been disrupted, even by Product 1.0. I wonder who among us will the first to be ambitious enough.
[Image (cc) by Visual Media]
I’d like to offer a few more predictions about Thomson Reuters Legal’s (TRL) latest version of Westlaw, known now only as Project Cobalt:
- Default search will not require Boolean connectors, however, the user will be able to switch to Boolean. I wouldn’t be surprised to see users run the same search with and without connectors just to compare the results. The user will expect the same results, and this is where I think education will be important. The results should be different with natural language processing, meaning better because it will undoubtedly be supported by semantic and sentiment analysis. This should yield better, more accurate search results.
- Search queries can be targeted to specific databases, but the menus will be consolidated. For example, I suspect law reviews will be collapsed under a single directory, and if the user wants to narrow the search to a specific journal, she will just type the name of it in the search box along with other search terms. Perhaps similar to Google’s “site” search feature, except more flexible.
- Search speeds will be significantly faster, but will probably still be throttled to maintain uniformity of user experience at all times of the day.
- Faceted search will finally make an appearance.
- Results Plus will be more tightly integrated into the search results. I don’t see this as a separate application any longer because the user needs to be able to appreciate, visually, the relationship and ranking of Results Plus hits to her primary search query. This may help drive in-platform upgrades.
- A “Westlaw Store?” I wouldn’t be surprise to see TRL borrow from Apple and allow users to set up a store account and to buy one-off access to books and databases in an effort to increase conversion to higher tier plans. How this could be accomplished without completely infuriating sales reps, I have no idea.
- I agree with Joe Hodnicki that I think we’ll see pricing packages similar to Microsoft, with TRL Solo, Small Law, Large Law, and Academic packages. I’m sure the matrix will be more complicated than this, and the need for publications outside of these packages might be satisfied by one off purchases through the Westlaw Store.
- Maintenance fees? Once you purchase a package, I wouldn’t be surprised to see a yearly maintenance fee as part of the offering. This is a common practice in the software industry, and can be adjusted by TRL every year, most likely based on your usage (which will still be heavily monitored).
- Free web results. The user will now have access to blog posts, law firm articles, etc. and will be able to click through to the them (meaning outside of TRL’s frame). I suspect these click throughs and leading content will be used as part of the sentiment and semantic analysis for the in-Westlaw results.
- More white space. I think we’ll see a cleaner, leaner look with minimal visual interference. The focus will be on search results.
[Image (cc) by jovike]
Three recent posts by Kevin Hunt over at Thomson Reuters Legal Current blog caught my attention this week because they reminded me about a video TR did about a year ago on the day in the life of a case. That video was sort of the first glimpse (for those of you who have never seen the process) into how much work (and people) go into putting a single case online. The latest videos give us yet another perspective on legal publishing through interviews with the people who are working to deliver analytical and other content to you. If you know nothing about the business, you might find these posts worthwhile, and I would encourage you to check them out:
- Creating the aha moment in legal research.
- Making legal research easier.
- Reference attorneys ready for your call.
[Image (cc) by duncan]
This year, Thomson Reuters Westlaw (which we’ll just refer to as Thomson Reuters Legal because, after all, that is what it will probably be rebranded as, with Tax & Accounting and Markets being the other brands) is poised to release a new version of Westlaw dubbed “Cobalt.” If you know what WALT means, then you’re old enough to appreciate how far Westlaw has come since the golden days of dial-up. And from what I can gather, Cobalt promises to be a huge step for CALR.
Here’s what we know about the new platform:
- Improved search engine.
- Easy to use. Think Googlesque. It is afterall, designed for the current generation and not “your dad’s army.”
- “Community insights.” I have no idea what this means, but it could hint towards social, which also might suggest an API.
- Somehow “optimizes” research workflow.
- Quick results. Think Loislaw maybe? Or perhaps they simply won’t throttle the return on results for the up-to 30,000 concurrent users so their experience is the same no matter what time of the day?
- “Nearly all high-value content exposed and used.” I’m assuming this means that you will now have access to most of TRL’s catalog, perhaps even backlisted items.
- Allows user to use common terms to search multiple databases. In other words, boolean search is history, and hello NLP.
- Learns from community. The example given is this: “if a particular research around bankruptcy results in people looking at a particular case and that case was really tangential, but not direct to the search, over time, as that case becomes more and more accessed, it will move higher and higher up in terms of the search criteria and selection that is provided to the user.” Source. Hopefully it will be able to learn in other ways.
Here’s what hasn’t been said.
- How are results presented?
- Are they still viewing Results Plus as an application, or is it incorporated into the results?
- Does my login still define my database collection?
- How much is the system relying on opencalais for improving semantic search? How much time will it take for the system to learn, and eventually when over the hearts of boolean diehards? For more on opencalais, check this video.
- What are “community insights”? Is there an expectation of customer engagement through the portal?
- Is everything still behind a walled garden, or does Cobalt have some windows?
- Will Cobalt leverage citations in briefs as a means for determining relevancy?
- Does Cobalt bring law, financial, and news together in one offering, as speculated by others?
- Does Cobalt come with a new pricing structure? One would think not because it is described as a “new search tool and design layout.” But then it is suggested that the platform will help the company “justify premium pricing,” presumably by quickly showing more and more relevant results. Not sure, but it’s a guess.
- Will there be an iPhone app? Or, is it optimized for mobility?
- In an effort to increase lawyers’ engagement with the site, will Cobalt be delivering video or other multimedia feeds alongside search results?
- Does Cobalt tap into the free web and return those results as well?
I can go on because there’s actually a great deal more TR has done on the Markets side that makes sense for Legal. Although it would appear the wait isn’t going to be that long, it sure would be nice to have a few more details now. For less information than what I’ve provided you but an opportunity to find out when Cobalt launches, go visit the pre-launch site here.