Thank you for sharing!

Your article was successfully shared with the contacts you provided.
Let’s face it. Practicing law and operating a law firm is growing more complex. Transactions cross more boundaries. Litigation increasingly means betting the company. Firms are transitioning from big to mega and from national to global. And the technology to support it all has now moved from the bowels of the back office to the belt loops of nearly every lawyer. How firms address this complexity and even turn it to an advantage will help define success and failure in an increasingly competitive legal landscape. Many firms, including most of the large global ones, have launched knowledge management programs to help them deal with this complexity. At Cleary, Gottlieb, for instance, we began an ambitious initiative four years ago to organize and enhance our knowledge management capabilities. Coupled with a technology upgrade that included the firmwide adoption of Lotus Notes and the development of a “virtual file room” of archived documents, e-mail, and images, our KM effort has yielded precedent collections of useful sample and standard documents, forums for shared know-how, deal-tracking databases, and practice-oriented “news centers.” The infrastructure is now in place to share work product between 800 lawyers in 12 offices around the world and to quickly push out information and expertise on significant legal developments. KM has also played a key role in various marketing, client relationship management, and associate development/training efforts at the firm. We have become victims of our own success, however. With dozens of intranet practice sites and Notes-based discussion forums and other databases to choose from, our lawyers are often baffled by where to begin and frustrated with the incompleteness of any one internal content source. For instance, a lawyer looking for information on a hot topic such as the new Securities and Exchange Commission Regulation G (resulting from Sarbanes-Oxley) would find lots of useful background information in our securities law forum, but might not realize that categorized samples of non-GAAP financial measures subject to Regulation G have been posted in a separate MD&A (Management’s Discussion and Analysis) database. Likewise, a thorough examination of recently implemented European Union competition law changes would require separate searches in our antitrust forum, in our European Union legislative developments forum, and in several of our local law forums. Compounding the problem is the almost perverse challenge of full-text searching our document management system, iManage, which because of configuration limitations requires separate log-ins to each office repository and the patience of Job to obtain and view the results from distant offices. Since results are not ranked intelligently and the document libraries are so full of “noise” documents like correspondence and interim drafts, lawyers rarely attempt to use iManage’s full-text searching capability to look for useful materials. “We need our own Google” was becoming the constant plea of our lawyers. Indeed, Google and other Web search engines have permanently raised user expectations about speed and comprehensiveness and have fundamentally changed search behavior. Even lawyers who grew up with the Boolean arcana of Lexis and Westlaw have grown intolerant of ands, ors, and withins. Web searching has taught them to expect remarkably prescient results within the first 20 hits from the simple input of a few keywords. Thus, our task was as clear as it was daunting: Find and implement a search engine capable of accessing and returning results from our iManage and Notes databases, intranet pages, and selected Internet pages. It better be fast, and it better be smart. THE SEARCH IS ON Our selection process commenced with the preparation and distribution of a detailed request for proposals to about 10 vendors, the initial group culled from online research ( www.searchengines.com is a good starting point) and discussions with colleagues at other firms. Among the leading vendors are Autonomy Inc. (Enterprise Search), IBM (Discovery Server), Recommind Inc. (MindServer), Triplehop Technologies (MatchPoint), and Verity (K2 Enterprise and Ultraseek). And, yes, there is even an enterprise version of Google available. The big legal search vendors have also gotten into the act with search engine offerings that leverage their proprietary legal taxonomies and content. West Group offers West km, and LexisNexis recently launched its Total Search product. We solicited several vendors already associated with the legal market (primarily those already packaged with a document management system), several that were well-known outside of legal circles, and several smaller vendors that offered something innovative or different. After initial presentations by most of the vendors, we winnowed the list down to three and asked them to index a sampling of about 160,000 Word documents, several Notes databases, and portions of our intranet and the SEC’s Web site. Documents from our European offices were later added to test multilingual capabilities. All three of our finalists had conceptual searching capabilities (i.e., the ability to return results based on the underlying concepts associated with a search term, not just the term itself). Large samplings of our securities and antitrust document collections and plenty of unrelated “noise” documents were used, but no effort was made to manually “train” or “tune” the concept searching capabilities of the tested systems. All three systems employed proprietary algorithms to intelligently rank results. A team of lawyers, paralegals, librarians, and knowledge managers conducted tests and provided “objective” feedback (first relevant hit, total relevant hits within the first 20 results, system response time, etc.) and subjective impressions. Key factors we considered in making our final decision included: • Integration with iManage. Most search vendors integrate out of the box with Exchange/Outlook and Notes, Web pages, and file systems, but integration with iManage and other document management systems such as DOCS is far less certain. • Effectiveness of the search algorithms. We found interesting variations in ability to rank relevant results and “intuited” results by returning documents that did not contain the exact search terms used. Given the inability to fine-tune the systems, we were encouraged by indications that the algorithms do work. (Bear in mind that litigation documents tend to be richer in conceptually unique content than are transactional documents.) • Look and feel. The user interfaces of all three engines were highly customizable. However, limitations in results display options, summarization, hit highlighting, and hit-to-hit navigation were observed. • Administrative functionality and ease of use. We found some significant differences in the flexibility and ease of use of the administration tools. Warning! Look behind the pretty end-user interface. • Search and indexing speed. We were able to compare performance between the three systems. While it’s difficult to extrapolate real-world performance from such a test, the ability to circumvent iManage system security overhead by storing security settings in the full-text index rather than in a separate database combined with the option to preview documents in HTML appears to positively affect performance and user satisfaction. • Vendor intangibles. Vendor stability, commitment to the legal market, and willingness to work with us were important considerations, as well. Price, while not exactly an intangible factor, was roughly the same between the vendors we looked at. The search engine market is full of pricing options that vary from dirt-cheap or even free to very expensive. The high-end features we were looking at (concept searching, categorization, and lots of flexibility) tend to come with a correspondingly high-end price tag. AND THE WINNER IS . . . We selected the MindServer system from Recommind Inc. MindServer can be configured with concept searching only, but we were equally impressed with the categorization feature that enables searching and results viewing by inferred topic or other category. We are also intrigued by the (as yet untested) potential of this system to infer entity information (names of parties, jurisdictions, case citations, etc.) and document types (memo, letter, agreement, brief, etc.) from the documents themselves. Such inferred metadata could then be fed back into the corresponding document profiles in iManage either automatically or through a work-flow process. Also, with MindServer, we can merge the relatively sparse profile information tracked in iManage with rich metadata collected in our KM deal databases. The prospect of combining intelligent full-text document searching with extensive and accurate matter metadata is truly exciting. We look forward to soon having a “universal” search system. Basic searching that encompasses all iManage libraries, Notes databases, internal Web pages, etc. will require no additional logins and can be initiated from anywhere within our intranet. Contextually narrowed searches will be just as easy. To limit the search scope to intellectual property sources, just go to the IP practice site on our intranet, type in the search terms in the search box at the top of the page, and click the “Search this Practice Only” option. For more flexibility use the “Advanced Search” option to customize the scope of the search by office, practice area, or data source. Search results will be returned by relevance, but the user will be able to apply numerous sorting and grouping options — for example, clustering the result set into folders grouped by topic or by client/matter. Only one version of an iManage document (the latest, unless a significantly more relevant earlier version exists) will be displayed, but access to all other versions will be only a click away. Another click will display all other documents for that matter. Implementation of MindServer is currently under way and will continue through the first half of 2004 as we bring up each of our 12 offices and multiple data sources. Although we are too early in our implementation to pass final judgment, we are very excited about this project and remain convinced that an effective universal search system will help us better manage the complexity of practicing law today. Brenton B. Miller is the director of knowledge management at Cleary, Gottlieb, Steen & Hamilton in New York.

This content has been archived. It is available through our partners, LexisNexis® and Bloomberg Law.

To view this content, please continue to their sites.

Not a Lexis Advance® Subscriber?
Subscribe Now

Not a Bloomberg Law Subscriber?
Subscribe Now

Why am I seeing this?

LexisNexis® and Bloomberg Law are third party online distributors of the broad collection of current and archived versions of ALM's legal news publications. LexisNexis® and Bloomberg Law customers are able to access and use ALM's content, including content from the National Law Journal, The American Lawyer, Legaltech News, The New York Law Journal, and Corporate Counsel, as well as other sources of legal information.

For questions call 1-877-256-2472 or contact us at [email protected]

Reprints & Licensing
Mentioned in a Law.com story?

License our industry-leading legal content to extend your thought leadership and build your brand.


ALM Legal Publication Newsletters

Sign Up Today and Never Miss Another Story.

As part of your digital membership, you can sign up for an unlimited number of a wide range of complimentary newsletters. Visit your My Account page to make your selections. Get the timely legal news and critical analysis you cannot afford to miss. Tailored just for you. In your inbox. Every day.

Copyright © 2021 ALM Media Properties, LLC. All Rights Reserved.