5. Problem: Our lawyers still waste a lot of time trying to find relevant work product (memos on the current topic, precedent agreements or interrogatories), even though we've spent a lot of money on a sophisticated search engine and major document management system.

Response: Improve the types and reliability of document metadata.

Discussion:

Document metadata are additional data saved with a document that give it contextual meaning, such as document type, title, author and date.

Document searching would be more efficient if a lawyer could limit his or her search to specific document types relevant to the inquiry, such as memoranda in the case of a research question or agreements in the case of a precedent search. However, document type is typically not a reliable piece of information, because it's gathered voluntarily from the document author when completing the profile and has not bearing to the author on any other action taken with the document. The information exists solely for the benefit of future searchers.

A different approach is to leverage off the lawyers' existing workflow. Most firms have word processing templates or macros to assist in creating the proper format and style for various document types. Assume a template is created for “Memo” in such a fashion that lawyers will find it easier to start their memo using this template. In that case, a small piece of customized software could be written that fills in the document type as “Memo” when the document is saved in the document management system.

The result would be much more reliable document type metadata.

Another approach, which is less accurate and more expensive than the previous one, could nevertheless be useful for improving the accuracy of document type metadata in an existing collection of documents. This approach employs auto-profiling software.  The software can improve searching by breaking a document population into document types, after first being “trained” on a population of known document types.

In addition, there may be other data within the firm that can be added to the document profile as metadata through linking of client/matter numbers. For example, if the firm has a matters and experience database (as described in Problem 3), information about the industry or type of matter can be added to the profile, enabling a search to be narrowed based on those criteria.