News You Can Use: A Robot Paralegal?

2052649a.jpg

The future is now! Dentons, LLP, the multinational firm boasting the largest number of lawyers in the world, apparently needs more of them — of the electro-mechanical sort. Dentons is financially backing a group of students from the University of Toronto who, as a class project, created an artificial intelligence to perform legal research. Code-named Ross, the AI uses IBM’s Jeopardy-nailing Watson to scour massive amounts of case law and legal docs in seconds to produce answers to legal questions. A Siri for lawyers, as the student-startup founders describe it.

Dentons is not new to embracing legal technology. It has its own legal tech lab, called Nextlaw Labs in California, where it apparently cooks up new legal technology projects. The student start-up, known as Ross Intelligence, Inc., will get access to the labs for their work, along with the financial investment from Dentons. Dentons, through Nextlaw, is also in the process of rolling out a cloud-based platform for lawyers, employing the IBM cloud and BlueMix Cloud Foundry technology. The idea is to create the foundation for developers to leverage and enter the legal tech market in order to create even better products for legal consumers to manage their work.

Right now, while the developers are refining and improving, Ross the Robot is focusing his intellectual powers on a database mostly comprised of U.S. bankruptcy law and “he” is being piloted not just at Dentons, but also at a few other elite firms in the US. His developers say that Ross is “learning” through action and responsive feedback – an exciting and frightening concept, to be sure. Pretty soon, I imagine, Ross will have the learned experience of a fifth year associate.

AI in the legal field is not unique to Ross – Microsoft’s Ventures Accelerator program is involved in another project that leverages AI to work through parts of complex contracts. RAVN Govern, powered by RAVN Systems’ Applied Cognitive Engine also uses AI to assesses  risk levels in contracts, and advise regarding the risk exposures before the contract is signed.

But no doubt, Dentons definitely has it all figured out. From the Denton’s release: “Technology is now and will continue to be a real differentiator in the legal profession,” explained Dan Jansen, CEO, NextLaw Labs. “The potential in companies like ROSS shows how the approach to solving client challenges is going to change. NextLaw Labs wants to be a part of transforming what is possible into a tangible offering in today’s legal market.” I, for one, welcome our new robotic paralegal overlords. Can I grab you a cup of coffee, Ross?

Advertisements

Cutting Edge Research Alternative Ravel Law

ravel-logo

Here’s another one for you: Ravel Law, a 2012 startup out of Stanford Law, promises radically easier, faster and more intuitive legal research option through its visual search engine. Apparently, Ravel (which promises to UNravel the law), is a collaboration between law, computer science and design. I can’t help but agree that the interfaces offered by the traditional legal database giants continue to fall WAY behind the curve on UI and ease of use.

Ravel uses a graphical visual interactive interface to display key information about cases. It is really something else entirely. I searched in the SCOTUS database the terms “insurance” and “mccarran ferguson” and got the result below:

insurance mccarran ferguson

 

When you hover over a circle, it takes you to a case in the results. You also can drop down in the list on the right to see the citing sources, click on them and get a new visual representation of how that case has been interacted with. You can quickly move along research trails using the visual interface, or scan in a more traditional model using the list on the right. I think it is a truly brilliant and novel approach to parsing out how court decisions interrelate – you can easily map out the history of cases and how and when search terms are cited in a manner completely different from traditional models.

Right now, the information being mined includes  cases and statutes from the federal circuits and SCOTUS back to about 1950. They are expanding as they can, but are somewhat limited to the extent that the underlying data is or is not presented in a machine-mineable format. The service is currently in beta and is free. Check it out and let me know what you think about it.

Casetext’s Wikipedia-Style Resource for Legal Research

casetext

An eBay for lawyers? How about a Wikipedia for legal research? They’re mashing up social media with the legal profession left and right these days. Casetext is an interesting, um, case, particularly here in the Studio where I am all about the free and cheap and the big Two Three have been a target of mine for years. Imagine. Making all that money off publicly available information.

Casetext is clearly intended to disrupt some of that. The hurdle that  free and cheap access to legal research materials has had to overcome is the value-add that comes from annotations and citation treatment. Lexis and Westlaw have certainly spent a lot of effort honing and promoting that value-add. Casetext’s angle is to get that value-add through crowd-sourced case annotations, much like Wikipedia does with its articles or Quora does with its Q/A format.

Casetext is the creation of two former law review heads from Stanford and Harvard.  Users of the service are encouraged to add tags and text to cases, link to other cases and generally provide similar data to that provided by the attorney editors at the big paid legal data companies. Contributors can provide  analysis of a document or of a paragraph within a document, link to their own articles or other related sources, add related cases and up-vote useful related sources. Contributions are licensed under a Creative Commons Attribution-ShareAlike license, permitting commercial use with proper attribution.

Users have to use real names, which hopefully encourages a higher standard of contribution than the anonymous commenting model. LIke any good social model, there is a reward system. Casetext uses reputation points, measuring a user’s contribution to Casetext. Gain points for adding content, for categorizing cases, for upvoting, and for receiving upvotes on content you add. Lose points for being downvoted and pay points to downvote others. Interesting system of checks and balances. There are some decent contributors on the site already, including a law professor who annotated a case he had argued to SCOTUS.

There are Quick Facts and a Document Wiki, essential information at a glance and and free form document summaries, respectively. Related cases are citing sources. The record includes oral arguments to SCOTUS. Create a PDF of a case with the two column format you may be very familiar with from the other guys. You can create a bookmark list of cases to read later, and even a Heatmap which highlights the most cited passages – dark blue means most cited. There is also a “copy with cite” feature – one of the features the Westlaw rep proudly touted to me when she was up-selling me on WestlawNext.

Its free to use right now, but is promising a paid Pro premium model. Right now, the big challenge is scope: the databases only include all U.S. Supreme Court cases, federal circuit court cases from Volume 1 of F.2d, federal district court cases published in F.Supp. and F.Supp.2d from 1980, and Delaware cases published in A., A.2d, and A.3d from Volume 30 of A. It was last updated on June 14, 2013. Hopefully it will open up to new jurisdictions soon. Quite frankly, I think this is a very exciting development, with a whole lot of promise if enough people play along.

Check out Casetext in action in the video below. What do you think? Would you contribute your expertise? Does the good of the many outweigh the good of the few or the one? Let me know.

Fastcase & William S. Hein Publishing – Like a Reese’s Cup

library-corner1

Just caught the news over a Slaw that my favorite cheap reasonably priced web-based legal research resource Fastcase has partnered with William S. Hein Publishing to offer inline hyperlinks to Hein subscribers that link to Fastcase  federal and state case law, while offering Fastcase users access to Hein’s historical state statutory data and law review collection in search results. Nice to see these well-respected resources partnering to offer more to subscribers. This is a benefit to those groups that is worth noting. Hein will get Fastcase’s primary coverage of SCOTUS opinions from 1754 to present, Federal Circuits 1924 to present, Board of Tax Appeals, Tax Court Memorandum Decisions, U.S. Customs Court, Board of Immigration Appeals 1996 to present, Federal District Courts 1924 to present, Federal Bankruptcy courts from volume one to present, as well as state case law from all 50 states, dating back from at least 1950. Fastcase will get Hein’s Law Journal Library, Seesion Laws Library, State Attorney General Reports and Opinions, and State Statutes: A Historical Archive. Hein gets Fastcase access at no charge and Fastcase gets Hein abstracts at no charge, but Fastcase subscribers will need the Hein subscription to get full access to Hein materials. These are the first secondary materials that Fastcase has sought to integrate, which is exciting news indeed. Anyone willing to take on the Big Two is o.k. in my book.

Tabulaw – A Drafting Tool With A Legal Bent


Whew! Gone for a few days as I dug myself out from a pile of work (the paying kind). Perhaps if I had a software tool that combined my legal research efforts and results with a simple, effective sharing and composition tool, I could whip out those opinion letters and research projects faster and more efficiently. Wait – you say there is something like that out there? Have you heard of Tabulaw?

Tabulaw is a web-based service that combines all of the tasks of researching and communicating a point of law into a simple copy / drag / drop interface. Tabulaw is the glue for your other resources and tasks – it appears that it will work with Google, Westlaw, Lexis/Nexis, and perhaps other databases (here’s hoping for Fastcase), allow you to copy and save sections of research, along with their appropriate citations, which are then available to you to drag and drop into your final document. I really like the idea of aggregating from different sources into one, citable “notebook” of content that can then be manipulated and shared. As far as the collaboration element, I am not sure how they intend to implement this – it would be uber cool to make these research folders open to multiple contributors, along with traditional social sharing or direct links to Scribd or Slideshare.

Tabulaw is in private beta and I don’t have an invite so I haven’t yet tested it myself. But you can bet that I signed up for the beta. If I get in, I will get back with more info on this promising tool. In the meantime, check out their promo video below.

Google Now Sending Scholar To You Via Email Alerts

Google Scholar made a big splash a few months back when it rolled out case law and law journals as an option within its Advanced Search function. People “buzzed” around the new service, wondering if it would be the big “Wexis” killer with its free format. While it has taken some pot-shots in the comparison tests, Scholar is still a more than viable means of securing legal information and is getting better by the day.

Google has taken another giant step closer to besting the big guys with its roll-out of free Alerts for Google Scholar results. Activating an Alert is as simple as running a search and following a few simple steps. After getting positive results, look for the envelope icon on the upper left corner of the page. Click the icon, select the number of entries you wish displayed and that is pretty much it. Alerts work for traditional Scholar results and, most importantly, legal cases and journals!

Alerts will bring you new material as it is entered into the database based on your search query and parameters. Advanced searching is available for Alerts to the same extent available in a regular search.  While you cannot limit your Alert to results from a smaller subset of sources, you can limit Alerts by author’s name or part of a name or from pre-built “collections” listed on the Advanced Search page.

While only available via email for now, Google likely won’t stop at that delivery source. As with traditional Google search alerts, I expect RSS feeds will be coming down the road.

I am unaware of how frequently Google updates its information or the length of the time span between information creation and entry into the Google Scholar system. One of the benefits of paid services is the quick turnaround time on data entry. Of course, speed is an issue with respect to discoverying new case law on a particular question. But, as a supplemental means of securing FREE notifications on legal searches as they happen on issues that do not require up-to-the-minute updates, it seems to me to be  impossible to beat Google’s combination of price and service!

Kudos, Google, to further tilting the legal research playing field in your direction!

Hat tip to Resource Shelf.

Cornell-Curated Collection of Legal Research/Writing Guides

Cornell University seal
Image via Wikipedia

If you are looking for some writings on legal research and writing, hit the jump to the Cornell Law Library’s list of legal research and writing guides (link here). The categories break down as follows:

  1. GENERAL WORKS ON LEGAL WRITING AND LEGAL STYLE
  2. GENERAL STYLE MANUALS
  3. CITATION GUIDES
  4. WORKS ON BRIEF WRITING AND ORAL ADVOCACY
  5. WORKS ON INSTRUMENT DRAFTING
  6. WORKS ON LEGAL RESEARCH

The categories above are links to the actual sections. While it is a “selected” and not a “comprehensive” list, there is more than enough to fill your office bookshelf!

Reblog this post [with Zemanta]

Research Tips From A Canadian Lawyer

Nada Khirdaji, a partner in the Research Department of Osler’s Toronto office, focuses her practice on legal research.  Sounds familiar to me. Ms. Khirdaji shares some of her suggestions for effective legal research in this article (link here) included in CCH’s January, 2010 Law Student Monthly column. She covers many excellent points – all focused on providing the researcher plumbing a new area of law with the world-view first and the finer points second. Her suggestions include: avoiding the case law databases as a first step and turning instead to general resources and texts; reading all cases cited in the notes; assuming application of a statute until proven otherwise, and performing a thorough review of the statute’s structure, location, table of contents and index; reviewing predecessor sections of statutes, related statutes from other jurisdictions and similar provisions in other statutes; avoiding journal articles for practical legal research; and, using firm-specific resources and consult ing colleagues.

Really, all very good advice. However, the suggestions seemed penned by an attorney from the 20th Century. Here in the 21st Century, there is another, important means for achieving a broad, world-view that Ms. Khirdaji omits. Consider looking at all available on-line resources as well, including the newly-fortified Google search, book searches, semantic search engines, and other curated legal databases and resources. Consider crafting your own custom search engines of pertinent governmental agency websites using Google Custom Search. Use the deep web search engines to find unconventional web documents. You never know what you may find.

Fastcase: An iPhone Lawyer's Fast Friend

Last week I mentioned it, this week I review it. I have downloaded and have been playing with Fastcase’s new, FREE, iPhone application that offers a portal to a database of caselaw and statutes at your iPhone-bearing fingertips. New to Fastcase? From their site:

Fastcase is a next-generation, Web-based legal research service, that puts the complete national law library on your desktop anywhere you have Internet access.  Fastcase’s smarter searching, sorting, and visualization tools help you find the best answers fast – and help you find documents you might have otherwise missed.

Fastcase reaches back to volume 1, page 1 of the major federal reporters and also includes primary caselaw from the 50 states and statutes as well as other materials  via its Web interface. Fastcase had already positioned itself as the affordable, legal research option, free to many via bar memberships. Fastcase is sure to blow that particular part of the field wide open with its free legal research iPhone application, the first of its kind to hit the App Store.

The iPhone application uses the same process as Fastcase’s web-based application – a natural language and relevancy weighted search function. But Boolean search works too. What’s more, there is a citation analysis tool integrated into the search results. No need for an existing Fastcase account: I quickly filled out a form within the app, providing my name, company, job title, and log-in information and achieved full access immediately. In their welcoming email, they include an offer for a free trial of their web product (aaahhh, THERE’s the marketing hook.)

Fastcase lists the features of its app as follows:

  • Free, searchable library of American cases and statutes
  • Keyword (Boolean), natural language, and citation search
  • Browse or search statutes
  • Most relevant results first
  • Customizable search results that you can sort five different ways
  • Search results automatically display number of citing cases
  • Jump right to most relevant paragraph of any case or statute
  • Integrated research history
  • Save favorite documents for use later
  • Case law is updated regularly

Not bad for a free application.

Of course, I have opened and closed the doors and hood, kicked the tires a couple of times and have run a few searches. In a word, it delivers.

I entered my search query:

set my jurisdiction and date parameters,

and let it run. Thirty seconds later, my results showed up.

The relevancy-weighted algorithm assigns a relevance score to the case from 1 (not relevant) to 100 (right on point). Authority check shows via an orange button indicating how many times the case has been cited overall and how many times it has been cited by other cases in the search results. You get more details on the citing cases when you press the button.

When you click on a particular case, it brings up the header with both the official and West Reporter citations (cool!), docket number, court, parties and officiating judges. There are arrow buttons to scroll through the results list, a save button to save a particular case (which you can also copy, paste and email to yourself if you want to print 😉 ) and a “most relevant” button which brings you to your search terms. The case text is easy to read and it hypertext-links you to other cases cited within the text.  Footnotes are displayed at the end of the case.

A burning question in my mind is how quickly Fastcase gets decisions loaded into its database. I haven’t queried Fastcase themselves and my research did not disclose a clear answer.

Another potential drawback for frequent users of the Big Two premium legal research products is the lack of the editorial treatment in these cases – no headnotes, keynumbers of introductory paragraph summarizing the case posture and holding. This editorial treatment is the value-add the Big Two lean heavily upon in keeping their prices at the top of the range. The lack of these features in Fastcase and, consequently, their app, may dissuade some from using this tool for more than the occasional quick look-up while on the go. But I do have to commend Fastcase’s relevancy algorithm – in an area of law that I am VERY familiar with, Fastcase quickly showed some of the key cases right at the top of the results list. Impressive showing for the first, free, legal research app for a mobile platform.

I give Fastcase’s app a hearty two-thumbs up for both effective implementation and the surefire “kick-in-the-pants” this free tool will bring to this somewhat archaic, top-heavy industry.

Run. Don’t walk.

Reblog this post [with Zemanta]

Westlaw Sings The Blues: Project Cobalt

Never able to let a mystery lie dormant, I spent some time this weekend digging into the dirt to find out about this West-branded next evolution in legal research. It’s code-named Cobalt and I managed to scare up a Thomson Reuters PowerPoint which hints at the features. It is fairly clear that Cobalt will offer a more Web 2.0 experience. Of course, its being billed as the best search engine for law and easy to use. It will promises “high velocity” results and  research workflow optimization. What interests me is a vague reference to “community insights.” Is West going to offer its own social aspects within the research framework?

Although it is not certain, there is a suggestion that the preferred search format will be natural language. Some are opining that this means no more Boolean. I will wait and see on that point before I assume the worst. I am hoping West is smarter than Bing in that regard. There is also a suggestion in the press that it will learn from the community – perhaps this learning hints at a semantic aspect (woohoooo!).

There isn’t a lot more to say at this point, but I will definitely keep my eyes peeled for more info. My inner cynic is moving aside a little to make room for a new hope that West will bring its service up to par.

Reblog this post [with Zemanta]