1 / 23

In favor of Isomorphic Intermediate Representation for Cross-Border eDiscovery

Global E-Discovery/E-Disclosure Workshop: A Pre-Conference Workshop at the 12th International Conference on Artificial Intelligence and Law . In favor of Isomorphic Intermediate Representation for Cross-Border eDiscovery. K. Krasnow Waterman kkw@LawTechIntersect.com June 8, 2009.

amelie
Download Presentation

In favor of Isomorphic Intermediate Representation for Cross-Border eDiscovery

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Global E-Discovery/E-Disclosure Workshop: A Pre-Conference Workshop at the 12th International Conference on Artificial Intelligence and Law In favor of Isomorphic Intermediate Representationfor Cross-Border eDiscovery K. Krasnow Waterman kkw@LawTechIntersect.com June 8, 2009

  2. Work is ongoing in multiple organizations and environments to create systems that can compute the correct legal response to a proposed data handling action. This presentation addresses considerations for successful communication of legal rules (law, regulation, contract, organizational policy) between lawyers and programmers. The goal is for programmers to be able to convert the rules to code which can be reasoned by machine (apply the rules to the action occurring in the system). Context

  3. Topics • Why intermediate representation? • Why isomorphic representation? • Applied to discovery rules • Comparing cross-border • Used in automated eDiscovery services • Conclusion

  4. Problem?

  5. What the computer scientist hears…. “The rule is that you have to produce the documents that the other party requests, if they made a formal production request, unless you have made objections (and they haven’t objected to them and you haven’t had a hearing or a ruling), and you had a pre-discovery conference and agreed on ground rules for production and how you’ll handle errors in terms of safe harbors for requesting and returning documents….blah, blah, blah…if there’s a one-eighth moon…blah, blah …party of the first part…” HUH?

  6. What the lawyer hears… “Did you mean: @forAll :PRODUCTION_REQUEST :REQUESTER, :RESPONDER, :RESPONSIVE_DOCUMENT, :DUPLICATE, :OBJECTION, . :ProductionPolicy a air:Policy; air:rule :RULE1. :RULE1 a air:BeliefRule; air:pattern { :REQUEST a k:RESPONSIVE _DOCUMENT; ; k:requestor :REQUESTER }; air:description (:REQUEST "is a request by an opposing party for documents in discovery " :REQUESTER); air:rule :RULE2, :RULE3.RULE4 :RULE2 a air:BeliefRule; air:pattern { :REQUESTER a k:OpposingParty; … math, math, more math …e=mc2 HUH?

  7. HELP! (We need a translator!)

  8. Need “intermediate” representation • Don’t speak the same language • Lawyers • Write concept-based, run-on text • Read left to right • Computer Scientists • Write short, logic-based expressions • Read top to bottom • Need a form both groups can readily assimilate

  9. It’s worked before! Public Domain. Courtesy of Wikipedia

  10. Need “isomorphic”* representation • Compound information differently • Lawyers • Add conditions, exceptions, and other compounding features • Computer Scientists • Compute the statements and represent the leanest form • Using a consistent structure is needed for lawyers to validate the expressions * Mapping into similar structure

  11. eDiscovery Example:

  12. Discovery • Assume search for relevant documents is complete • Can a system identify all the documents which the responder may exclude, or to which the responder may object, on grounds other than relevance? • Requires a system that can represent all the legal rules needed (about objections and exclusions) and reason over them. • Need to represent rules in a manner that lawyers can write and/or validate. • Need to represent rules in a manner from which programmers can readily create code.

  13. Simplest example of intermediate isomorphic representation

  14. UK rule re: multiple copies • “Disclosure of copies • 31.9 • (1) A party need not disclose more than one copy of a document. • (2) A copy of a document that contains a modification, obliteration or other marking or feature • (a) on which a party intends to rely; or • (b) which adversely affects his own case or another party’s case or supports another party’s case; • shall be treated as a separate document.”

  15. Lawyers validate by reading across Computer scientists/system developers read down: identify ontologies, range of variable values, representation issues, needed equivalencies Note: For actual use, requires at minimum significantly more abstraction of “information” and expression of the logical relationship of each rule to any other rule (e.g., condition, exception).

  16. Invert (and rearrange) for programmers to see the rule in a familiar vertical form Produce_Rule_1 If {Information: Document Information_Condition: Duplicate; Jurisdiction: UK} Then {Actor: Litigant; Rule: Not_Required; Use_Event: Produce}. Produce_Rule_2 If {Actor: Litigant; Rule: Not_Required; Use_Event: Produce} Then {Print: Document_ID In File: Documents_Not_Produced}. If Then Comment

  17. Cross-border • Can a system identify the differences between the discovery rules of two jurisdictions? • Can a system recognize and apply the discovery rules of the correct jurisdiction? • Requires a system that can represent legal rules and reason over them. • Need to represent rules in a manner that lawyers can write and/or validate. • Need to represent rules in a manner from which programmers can readily create code.

  18. US rule re: multiple copies • “Rule 26. Duty to Disclose; General Provisions Governing Discovery • (b) • (2) • (C) When Required. On motion or on its own, the court must limit the frequency or extent of discovery otherwise allowed by these rules or by local rule if it determines that: • (i) the discovery sought is unreasonably cumulative or duplicative, or can be obtained from some other source that is more convenient, less burdensome, or less expensive;”

  19. Cross-border comparison PATTERN MATCH (NOT RELEVANT) DIFF

  20. Better abstraction provides more granularity – clearer result NEW VALUE BETTER IDENTIFIES THE DIFFERENCE

  21. Applying Correct Law • In order for a system to determine which jurisdiction’s law applies • Need an isomorphic representation of all relevant rules for resolving conflict of laws • Will likely require such values as • Current location of data • Location of requestor • Location of data collection • Conditions under which data collected • Existing related agreements between nations

  22. Conclusion • Table, per se, is not important • What is important is to obtain lawyer trust by allowing them to write (or see and validate) for themselves what rules the system is using. • What is generally not important is for lawyers to see how the system is reasoning. • Goal is any structure that • Can produce intermediate representation • Readily understood by both parties • Consistent in structure • One-for-one representation • To which logic rules can be applied • Translation • Equivalencies • Conflict resolution • From which programmers can readily produce code

  23. References • Please see related paper: • Isomorphic Intermediate Representations Are Needed To Support Cross-Border eDiscovery and Digital Evidence Systems • http://www.law.pitt.edu/DESI3_Workshop/Papers/DESI_III.KKW.05809.FINAL.pdf • Citing: • [1] Daley, MJ & Rashbaum, KN, eds., A Practical Guide to Navigating the Competing Currents of International Data Privacy and eDiscovery, The Sedona Conference (Public Comment Version, 2008). • [2] Bench-Capon, TJM & Coenen, FP, Isomorphism and Legal Knowledge Based Systems, Artificial Intelligence and Law, Vol 1, No 1, pp65-86 (1992). • [3] Karpf, Jorgen, Quality assurance of Legal Expert Systems, Jurimatics No 8, Copenhagen Business School, 1989. • [4] Weitzner, Abelson, Berners-Lee, Hanson, Hendler, Kagal, McGuinness, Sussman, Waterman, Transparent Accountable Data Mining: New Strategies for Privacy Protection,; MIT CSAIL Technical Report MIT-CSAIL-TR-2006-007 [DSpace handle] (27 January 2006). • [5] Kagal, Hanson, & Weitzner, Integrated Policy Explanations via Dependency Tracking, IEEE Policy 2008 (cited in Acknowledgements). • [6] Hammar, PK, of PKH Enterprises, MD co-lead.

More Related