by “how” I mean, is it used via a browser like I use Google? Or do people use elaborate specialized client apps specifically intended for this (presumably fairly specialized) task?
In terms the user interface, is it any better / more sophisticated than the basic boolean keyword search google experience? Are there more creative ways to filter results across repeated searches, attach notes to documents etc?
Going a bit further into the IMHO territory, is the Lexis Nexis (or any other legal-related) search activity a sufficiently important and valuable activity for lawyers to care about making it more efficient, powerful etc? E.g. as a programmer I get to think a lot about quality of search for programming related docs using google, and I can’t help but notice that much of the time little time spent on this, except for those few times when it matters a great deal. Which suggests that if we want to improve search for programmers we should either research complicated solutions for the infrequent complicated situations or, just do what Google does and ignore the whole issue as commercially useless (even if practically relevant even to their own employees whose hourly salary is high enough that keeping them searching for too long even on infrequent basis is not good).
So how about the lawyers? Do they spend enough time in Lexis Nexis to care about the “Lexis Nexis experience”?
You can access Lexis Nexis only if you subscribe to it. It is a website. Many lawyers find it very useful. When I worked for the SSA, I was given a subscription to it. (Since my retirement, I haven’t renewed it.)
The legal department at the bank I used to work for had Lexis-Nexis, and they wouldn’t live without it for an hour if my memories of panicked phone calls are any indication. Inability of a legal dept. staffer to access Lexis was one of my very few drop-all-other-issues-type Top Priority calls. About the only thing that would have trumped it would have been a fire in the server room or the zombie apocalypse.
Back then (1997-ish), it was accessed via a dedicated client-server application installed on the PC with hardware-key control (IIRC, a dongle plugged into one of the PC’s serial ports) and a dial-up connection to the server. A lot has changed since then, though.
One of the big issues is that if you’re at a smaller firm that doesn’t have unlimited access, is that you get charged per search so inefficient searching can get seriously expensive. Even if you do have access, lawyers’ time is expensive and so an efficient searcher can save quite a lot of money. Legal research also isn’t the scavenger hunt regular research can be-- it’s usually not a matter of finding a particular nugget of information, but sifting through all the extremely similar data to find the particular data that’s relevant to the jurisdiction, situation, etc.
I don’t know about Lexis, but the other legal database recently launched a new search engine/algorithm that is apparently so powerful that a friend of mine who goes to law school says its considered cheating in some classes.
It also, or at least it did, back when I worked for a state agency PR office. A terminal was in our legal section, but we used it to download news stories from almost all U.S. newspapers. It was not easy to use, but being the resident geek, I was the only non-lawyer who figured out the system.
Lexis is far more complicated and specific than google. Some of the basic search options in addition to the basic boolean AND / NOT include wildcards (neglig! returns both negligence and negligent) proximity searches (search for “negligence” and “comparative” within # of words of each other, or in the same sentence, or in the same paragraph). ATLEAST (return only results that use the search term at least # of times). Further refinements are available by date, by jurisdiction, by court type, and “search within” your first set of results.
There are also various ways to browse statutes and written materials. You can also search directly by citation, or by party names and/or court docket number within a specific jurisdiction. Once you’re in a case you can find prior and subsequent history (what happened in lower and upper courts, if anything did) sometimes read pleadings, and find out if any other cases have cited the case you’re looking at. If a case has been overturned, it is automatically red-flagged. If someone has disagreed with it, it is yellow-flagged.
Similarly, statutes are generally annotated - joined with cases that have discussed or applied that specific law.
Learning to use Lexis, and its comeptitor, WestLaw, is a fairly non-obvious task. law schools offer classes in using the functions, and student reps are generally available on school grounds during all school hours.
Both Lexis and Westlaw changed their websites significantly this year.
(using the pharmaceutical industry because it is familiar to me)
Suppose a scientist at Bristol-Myers Squibb is interested in searching for preclinical study reports for in-house drugs that caused liver damage in dogs.
They could say “Find preclinical study reports for a corporate drug that causes liver damage in dogs”
This search could be applied to an ontology which would then find study reports that contain those concepts.
For example, a particular study might mention “beagle” and “liver necrosis” and “BMS-013241”
That report would be included because “beagle” matches “dog” in the ontology, and “liver necrosis” matches “liver damage”, and “BMS-013241” matches some internal identifier in a dictionary of known BMS drugs.
This is a kind of searching that is fairly narrow scope to an industry (the ontologies need to be tailored to the business domain), but it goes way beyond what Google provides in many cases.
I’m afraid I don’t totally understand your question (sorry!).
Both Lexis and Westlaw are tremendously stupid at natural language searching, although they offer the option, the results are useless.
It is also important to understand that although they both are served through the internet they don’t search the internet. They search their own proprietary, insanely extensive databases and indices. I believe Westlaw has around 40,000 indicies and databases that underlie their search function, all maintained by them. West came up with its system of classifying cases by topic in the 1800s. They still use essentially the same system today!
So if no index exists of all “corporate drugs” you can’t use a Lexis/Westlaw search that targets that general category of item in the way you’re describing.
Assuming that such an index or database does exist,
The equivalent search would be:
[select database containing corporate drugs]
dog /p necro! OR “liver damage”
(search for “dog” in the same paragraph as any word beginning with “necro” OR the phrase “liver damage”
Once you find a result you like, you can also use either the “more like this” function or the Headnotes/Keynote function (which is the indexing system, which categorizes by legal issue, as I mentioned above) or the “view cases that cite this selection” function to find more, related items.
The Ontology bit is where a search tool would magically connect Beagle with Dog without you having to tell it so. They can be created for diverse business domains. For example, it would be cool to have an ontology of music, with genres, bands, artists, albums, and the like, so you could search for “Jazz Musicians” and get hits on “Wes Montgomery”
A friend put together an ontology for beer terminology to demonstrate an indexing product.
This kind of stuff, along with proximity (e.g. “within 5 words of”) is cool stuff, and I always wonder how many industries leverage natural language processing to go beyond simple full-text searching.
everybody, thanks for replying with interesting info.
So overall it sounds like the answers to my OP questions are:
as of 2011 legal search seems to be done mostly through web interface in the browser, although in the past special purpose client may have been used (at least this is implied upthread, but not sure if clearly confirmed)
yes, lawyers really do spend lots of time searching and care about this issue. Quite unlike the google geeks who don’t search all that much and, perhaps for that reason, don’t knock themselves out inventing superior front-end search user interfaces.
The above suggests that if the process of legal searching and associated note-taking could be “significantly” improved purely on the client side, without messing with the back-end, such improvements could be implemented by an ISV or academic researchers (i.e. by an entity separate from Lexis Nexis) essentially as a browser plugin / browser automation app that works with the Lexis-Nexis website. So the basic “browser displaying a search website + MS Word for taking notes” system could evolve into “search query and results history manager + OneNote v2.0 specialized for the format of legal documents and associated notes”
A possible next step in this system’s evolution might be incorporation of human virtual assistants tasked with performing (for a lower hourly wage than the main user) some of the operations involved in searching and evaluating the documents. This might further be subdivided into a high trust case of managing paralegals employed by the same law firm (i.e. the only “virtual” aspect here is that the results of their work are accessed through computer interface) and the low trust case of outside remote service providers.
Incidentally, this may be thought of as one possible answer to the question I have asked in a previous thread of mine are there web frontend enterprise systems sufficiently uncustomizable to benefit from UI automation? . In this case the lack of user interface customizability is due to the system being designed and run by a big outside company that imposes its own ideas, or lack of them, about how the interface should work.
That’s not the the only obstacle. The other one is the almost indescribable technological conservatism of the legal profession. I was the ONLY person in my entire office who wanted to switch to WestlawNext (the newest interface offering). Some about had a panic attack at the thought. In fact, I believe that rather than treating the new interface as a premium service, they had to discount it to get people to try it. (and not because it was confusing – it was very obviously simpler to use). I’m sure somewhere there are partners grumbling because they shutdown Westmate (one of the proprietary clients) in 2007. That was 10 years after they started offering a web-based interface.
As a long time Nexis user (Lexis is the law half, Nexis is the business half), I too have wonder why Google doesn’t use a ‘w/5’ type function like Lexis and Nexis use to say “only bring up articles where this word appears within five words of that word” because it would really help narrow searches much faster. At my previous two companies, I used Nexis for doing market research on technologies and for developing marketing plans. The wonderful thing about Nexis is that someone somewhere has written an article on how big the widget market is going to be for your esoteric technology or how it can be applied to the X,Y, and Z industry that really helps you make a business case, particular when writing grants. While that information is often on the Web as well, you have to sort through a lot of crap to get there because some asshole always has included your word in their porn site somewhere to drive traffic there, which I find really annoying. Then again, Nexis is pretty expensive as I recall. Better to head over to the university and use it for free on their computers.
Another thing to add is that at our Law Firm we had our own database of forms, filings, briefs, etc. tied into our Lexis Nexis system, meaning that when we used Lexis it searched our own (proprietary) data along with the Lexis database, and all of the documents were mapped appropriately (our to theirs, theirs to ours). It was fairly amazing.
eta: I also got occasionally bitched at and showed the costs of my searches, which could top $1,000/day - although that was at ‘sticker’ price and we didn’t pay nealry that much - not sure why they didn’t show us the actual cost, would have been way more useful.
They give you free access in law school. You know, just like that old PSA “what, I just give it away for free?” Out in the real world, searches are expensive and billable to a client. Many horror stories of associates (especially summer associates) accidentally running up huge bills.
I always wondered why Google didn’t partner up with Lexis or West to revamp their index. Imagine finding cases based on how often that case is cited. Not perfect, but Google’s engine would vastly improve the experience and usefulness. If either company was first to adopt, they’d pull amazing amounts of business away from their competitor.
Are you suggesting the creation of a software intermediary between the user and the Lexis-Nexis interface? I suspect that the Lexis-Nexis end user agreement would prohibit something like that.
Larger law firms do have support staff with lower billing rates to do routine work. The question of what work is done by higher billers and what work is done by lower billers can be a contentious one. Generally, law firms like to be able to bill in-house rather than subcontract work out to separate support services.
Note that Lexis-Nexis and Westlaw do offer pretty good remote help desk services when you find yourself stuck.
