Communication | Business communications » Zeleznikow-Bellucci - The Role of Principles of Justice in Building Mediation Decision Support Systems

Source: http://www.doksinet The role of principles of justice in building mediation decision support systems John Zeleznikow Faculty of Law, University of Edinburgh, Old College, South Bridge, Edinburgh, EH8 9YL, Scotland, UK john.zeleznikow@edacuk Emilia Bellucci Department of Computer Science, La Trobe University, Bundoora, Victoria Australia, 3086 bellucci@cs.latrobeeduau Abstract In this paper we discuss Artificial Intelligence techniques for building Negotiation Support Systems. We take our examples from Australian Family Law, where there can often be a conflict between negotiation and justice. Keywords: Decision Support Systems, Litigation, Negotiation 1. Introduction (Ross1980) states The principal institution of the law is not trial; it is settlement out of court. So what influence does judicial decision making have over the outcome of negotiated settlements? The answer is a major one, since judicial decisions serve as the very basis from which negotiations commence

(Williams 1983). Litigation can be damaging to both parties in a dispute. It is a zero-sum game; in that what one party wins the other loses. 1 Mediation can strive to reduce hostility between the parties, to fashion an agreement about tasks each party is willing to assume and to reach agreement on methods for ensuring certain tasks have been carried out. It can lead to a win-win result. 2 (Chung et al 1997) stress that although dispute resolution is a human problem, computers are already at the bargaining table to transform the negotiation process. The Harvard Negotiation Project (Fisher and Ury 1981) introduced the concept of principled negotiation which advocates separating the problem from the people. Fundamental to the concept of principled negotiation is the notion of Know your best alternative to a negotiated agreement (BATNA) -the reason you negotiate with someone is to produce better results than would otherwise occur. If you are unaware of what results you could obtain if the

negotiations are unsuccessful, you run the risk of: (1) Entering into an agreement that you would be better off rejecting; OR (2) Rejecting an agreement you would be better off entering into. (Sycara 1998) notes that in developing real world negotiation support systems one must assume bounded rationality and the presence of incomplete information. Our model of legal negotiation assumes that all actors behave 1 It is actually worse than a zero-sum game and indeed can often lead to a lose-lose result. This is because of the large legal fees arising from litigation. 2 For example if both parties value the list of items in dispute, it is not uncommon (as long as they do not value the items in an identical manner) for each party to receive 70% of their requested points. Source: http://www.doksinet rationally. The model is predicated on economic bases, that is, it assumes that the protagonists act in their own economic best interests. In this paper we discuss various approaches we

have developed to building negotiation support systems. We confine ourselves to discussing two-party disputes. Techniques we have developed include: i) Template based systems; ii) Knowledge-based systems that advise upon BATNAs; iii) Hybrid rule-based/case-based systems; iv) Game theory v) Integrating game theory and knowledge based systems to advise upon trade-offs We claim that the techniques we have developed can be generalized to other domains. Whilst developing negotiation support systems, we have noticed that a fundamental conflict arises – is our system concerned with supporting mediation or providing justice? We illustrate this issue with examples from the domain of Australian Family Law. 2. Negotiation Support Systems in Australian Family Law (Katsh and Rifkin 2001) state that compared to litigation, Alternative Dispute Resolution has the following advantages: a) Lower cost; b) Greater speed; c) More flexibility in outcomes; d) Less adversarial; e) More informal;

f) Solution rather than blame-oriented; g) Private To avoid the risks of extra costs and an unfavourable outcome, disputants often prefer to negotiate rather than litigate. Whilst investigating how disputants evaluate the risks of litigation researchers are faced with a basic hurdle - outcomes are often, indeed usually, kept secret. If the case is litigated, it could be used as a precedent for future cases, which may be a disincentive for one or more of the litigants (Goldring 1976). Publicity of cases and the norms resulting from cases makes the public aware of the changing attitudes towards legal issues. 3 The adjudication decision not only leads to the resolution of the dispute between the parties, but it also provides norms for changing community standards (Eisenberg 1976). This latter facet is lost in negotiated settlements The secrecy behind negotiated settlements is one of the reasons for the paucity of published material on legal decision support systems dealing with risk.

WIRE IQ (Wire Intelligent Quantum) is an Internet delivered decision support system which allows lawyers, insurers and re-insurers access to up-to-the minute quantitative analysis of current claims settlement values for a wide range of personal injuries (Douglas and Toulson 1999). Douglas and Toulson (1999) state that analysis and price discovery of tort in un-settled personal injury claims has been conducted using rule-based systems. In such systems, the details of the claim (injury type, claimant’s age, sex, earnings, etc) are entered into the system. The system then applies predefined rules to determine the settlement value of the claim WIRE IQ uses a database with thousands of records of settled claims and court wards for a range of personal injury claims. It then provides the following analysis services based on the data: trend analysis, comparative analysis, precedent search and forecasts. The forecasts are performed using neural networks 3 In common law countries, changing

community values towards issues such as abortion, euthanasia and rape within marriage have been enacted in the legal system through landmark precedents, rather than parliamentary legislation. Source: http://www.doksinet In our desire to construct decision support systems to support legal negotiation, we realise how much the building of such systems depends upon the domain context. We chose as our domain to be modeled Australian family disputes In most legal conflicts, once a settlement is reached the parties to the settlement are not required to have an on-going relationship. This is not the case in Australian Family Law Family Law (Ingleby 1993) varies from other legal domains in that in general: 1. There are no winners or losers - save for exceptional circumstances, following a divorce both parents receive a portion of the property and have defined access to any children. 2. Parties to a family law case often need to communicate after the litigation has concluded. Hence the

Family Court encourages negotiation rather than litigation. The overriding principle in Australian Family Law is the paramount interests of the children. Many men’s rights groups have claimed Australian Family Law is feminist. Whilst there is no basis for such claims in any legislation, the overriding fact is that following divorce, the place of primary residence for most children is with their mother. With respect to the residency of children of divorced Australian couples, in 90% of non-litigated cases and 60% of litigated cases, the primary residence of all children of the marriage is in their mother’s home. Given this fact, and that the Family Court of Australia is obliged to place the financial interests of children first, it is inevitable that mothers will receive a greater share of marital property than would childless women. This is particularly the case in marriages that have minimal financial resources. We have developed numerous systems to support negotiation in

Australian Family Law. We claim all of the techniques developed can be used in other negotiation domains. 2.1 Deus – a template-based negotiation support system Our first attempt at building negotiation support systems was to build a template-based system, DEUS (Zeleznikow et al 1995). In building DEUS, we developed a model of family law property negotiation, which relies upon building a goal for each of the litigants, with the goals being supported by their beliefs. Goals can only take real number values, because in simplifying the model it is assumed that the goal of each party is a monetary figure. Beliefs, which support the goals, are expressed in natural language. In the system, which has been implemented using this model, goals are used to indicate the differences between the parties at a given time. The beliefs provided are used to support the goals. The model calculates the agreement and disagreement between the litigants’ beliefs at any given time. The agreement and

disagreement are only in relation to the beliefs and hence do not resolve the negotiation. In order to reach a negotiated settlement, it is essential to reduce the difference between the goals to nil. Having defined the model, it was implemented into DEUS. The system supports the negotiation process by representing the goals and beliefs of the opposing parties to a property conflict arising from a divorce application. It helps mediators understand what issues are in dispute and the extent of the dispute over these issues. Whilst DEUS does not perform belief revision or indeed does not have any intelligent functions, it does inform disputants as to their level of disagreement. DEUS is not concerned with the principles of justice Its goal is to indicate the range of issues in dispute. 2.2 Split-Up In dealing with the distribution of matrimonial property following divorce in Australia, (Stranieri et al 1999) determined that the task of determining what property a Family Court of

Australia judge may distribute was determined to be rule-based 4. The section of the Act dealing with the percentage of the matrimonial property each partner receives is highly discretionary. This is because the Family Law Act (1975) lists a number of factors to be considered for a percentage split determination yet provides no guidance on the relative significance of each factor or on how they are to be combined. Ascertaining knowledge about how a judge weighs and combine factors is difficult, in that a guessed numerical weight is unlikely to represent the actual weight of the factor in the context of a large number of interdependent factors. 4 A (crisp) rule is of the the form IF <condition(s)> THEN <action>. An example of a rule is if you drink and drive then you lose your licence. Source: http://www.doksinet (Fayyad et al. 1996) define knowledge discovery in databases (KDD) as the non trivial process of identifying valid, novel, potentially useful understandable

patterns in data. The aim of the approach used in developing Split Up was to identify relevant factors in family law with experts and then assemble a dataset of values on these factors from past cases that can be fed to machine learning programs such as neural networks 5. Percent time he worked full time Source of her negative contribution Extent of her negative contribution Source of his negative contribution Extent of his negative contribution Percent time he worked part time Percent time she worked part time Percent time she worked full time Comparative salary rate Resource name Resource type Dollar value from husnband Husand $ Number of years ago Relative effort by way salary Dollar value from wife Wife $ Number of years ago Relative contributions toward cash, assets or businesses Relative negative : Relative main Length of marriage Relative labor contribution Relative superannuation Relative homemaker Number of years of marriage Time apart during cohab Contributions

of husband relative to wife Number of years of cohabitation Common pool Name of dependent Relative household duties Age of dependent Care arrangement Relative child rearing Special needs His obligation to each dependent : His obligation to each dependent His salary amount Current realisable value Resource name His investment resources Resource type $ annual inome $ asset value His financial assistance from other sources : Demand for his skills A His obligation to all dependents B Her obligation to all dependents His financial resources A Future needs of husband relative to wife C His superannuation Payout date His age group His health His salary resources B Level of wealth He currently works He can work His education His financial responsibility to children His extraordinary financial expenses Her financial resources C Her age group Her health His age group His health His employment prospects Employment is available His work experience Figure 1. Tree

of actual arguments for Split Up 5 A neural network receives its name from the fact that it resembles a nervous system in the brain. It consists of many self-adjusting processing elements cooperating in a densely interconnected network. Each processing element generates a single output signal which is transmitted to the other processing elements. The output signal of a processing element depends on the inputs to the processing element: each input is gated by a weighting factor that determines the amount of influence that the input will have on the output. The strength of the weighting factors is adjusted autonomously by the processing element as data is processed. Percentage of pool assets to the husband Source: http://www.doksinet inference mechanism neural network DATA H has contributed X relative to the wife H has Y resources relative to the wife The marriage is of Z wealth Why Data is Statute makes Statute makes relevant this relevant this relevant Backing Section 79(4)

Section 75(2) CLAIM Husband is likely to be awarded P percent of assets Precedent cases Network trained with appropriate examples Network is trained with a proven learning rule: BP Why inference is appropriate Lee Steere, Brown 103 unreported cases Studies cited by Haykin Backing RELEV ANCE WARRANTS INFERENCE WARRANTS Figure 2. Toulmin argument structure for one of the Split Up argument In this way, the manner that judges weighed factors in past cases could be learnt without the need to advance rules. This approach was inspired by the jurisprudence movement known as legal realism. For legal realists exemplified by (Llewellyn 1962), rules and principles may be invoked after a decision has been reached in order to ensure that a decision is just, moral and legally correct. Rules and principles are invoked to explain a decision but there is no need to assume they are used to reach the decision. Ninety-four variables were identified as relevant for a determination in

consultation with experts. The way the factors combine was not elicited from experts as rules or complex formulas. Rather, values on the 94 variables were to be extracted from cases previously decided, so that a neural network could learn to mimic the way in which judges had combined variables. In the Split Up system, the relevant variables were structured as separate arguments following the argument structure advanced by (Toulmin 1958). Toulmin concluded that all arguments, regardless of the domain, have a structure which consists of six basic invariants: claim, data, modality, rebuttal, warrant and backing. Every argument makes an assertion based on some data. The assertion of an argument stands as the claim of the argument. Knowing the data and the claim does not necessarily convince us that the claim follows from the data A mechanism is required to act as a justification for the claim. This justification is known as the warrant The backing supports the warrant and in a legal

argument is typically a reference to a statute or a precedent case. The rebuttal component specifies an exception or condition that obviates the claim. In the current version of Split-Up only rules and neural networks are used. Previously, (Zeleznikow et al 1996) considered the use of rule induction 6, whilst (Stranieri et al 2000) considered the use of association rules 7 for discovering knowledge about the distribution of marital property. Split-Up can be used to determine one’s BATNA for a negotiation. It first shows both litigants what they would be expected to be awarded by a court if their relative claims were accepted. It gives them relevant advice as to what 6 A rule induction system is given examples of a problem where the outcome is known. When it has been given several examples, the rule induction system can create rules that are true from the example cases. The rules can then be used to assess other cases where the outcome is not known. 7 An association rule is a rule

that is not crisp. Non-crisp rules have both an associated support and an associated confidence. In the drink driving example, the support of the rule indicates what percentage of all drivers tested have been drinking, whilst the confidence of the rule indicates what percentage of drivers who have been drinking and driving lose their licence. A crisp rule has confidence 100% Source: http://www.doksinet would happen if some or all of their claims were rejected. Users are then able to have dialogues with the system to explore hypothetical situations to establish clear ideas about the strengths and weaknesses of their claims. Suppose the disputants goals are entered into the system to determine the asset distributions for both W and H in a hypothetical example. For the example taken from (Bellucci and Zeleznikow 2001), the Split-Up system provided the following answers as to the percentages of the marital assets received by each party: W’s% H’s % Given one accepts W’s beliefs

65 35 Given one accepts H’s beliefs 42 58 Given one accepts H’s beliefs but gives W custody of the children 60 40 Figure 3. Use of Split-Up to provide negotiation advice Clearly custody of the children is very significant in determining the husband’s property distribution. If he were unlikely to win custody of the children, the husband would be well advised to accept 40% of the common pool (otherwise he would also risk paying large legal fees and having on-going conflict). Whilst Split-Up is a decision support rather than negotiation support system, it does provide disputants with their respective BATNAs and hence provides an important starting point for negotiations. However, more is required of negotiation support systems. Namely they should model the structure of an argument and also provide useful advice on how to sequence the negotiation and propose solutions. Because Split-Up models knowledge about Australian Family Law Property Distribution, it is greatly concerned

with the role of justice. Indeed, Split-Up is concerned with providing appropriate legal advice, rather than supporting mediation. The very fact that Split-Up supports mediation is a useful, secondary goal 2.3 Family Winner and game theory based approaches for developing negotiation support systems (Jennings et al 2001) developed a generic framework for classifying and viewing automated negotiations. This framework was then used to analyse the three main methods of approach that have been adopted to automated negotiation, namely: 1) Game theory 2) Heuristics 3) Argumentation based approaches. (Bellucci and Zeleznikow 2001) have used all three techniques in building negotiation support systems. Family Negotiator (Bellucci and Zeleznikow 1997) is a hybrid rule-based and case-based system which attempts to model Australian family law. The system models the different stages of negotiation (according to Principled Negotiation Theory) by asking individuals for their positions and

reasons behind these. Game theoretic techniques and decision theory were the basis for AdjustedWinner (Bellucci and Zeleznikow 1998), which implemented the procedure of (Brams and Taylor 1996). AdjustedWinner is a point allocation procedure that distributes items or issues to people on the premise of whoever values the item or issue more. The two players are required to explicitly indicate how much they value each of the different issues by distributing 100 points across the range of issues in dispute. The Adjusted Winner paradigm is a fair and equitable procedure At the end of allocation of assets, each party accrues the same number of points. It often leads to a win-win situation Although the system suggests a suitable allocation of items or issues, it is up to the human negotiators to finalise the agreement acceptable to both parties. Arising from our work on the AdjustedWinner algorithm, we noted that 1) The more issues and sub-issues in dispute, the easier it is to form trade-offs

and hence reach a negotiated agreement; 2) We choose as the first issue to resolve the issue on which the disputants are furthest apart - one wants it greatly, the other considerably less so. Source: http://www.doksinet Instead of using points as in AdjustedWinner, we use influence diagrams in Family Winner. We then reformulate the influence diagrams with the aim of eventually reaching equality. Family Winner (Bellucci and Zeleznikow 2001) uses both game theory and heuristics. It supports the process of negotiation by introducing importance values to indicate the degree to which each party desires to be awarded the issue being considered. The system uses this information to form trade-off rules The trade-off rules are used to allocate issues according to the logrolling strategy. The system makes this analysis by transforming user input into trade-off values, used directly on trade-off maps, which show the effect of an issue’s allocation on all unallocated issues. Users of the

Family Winner system enter information such as the issues disputed, indications of their importance to the respective parties and how the issues relate to each other. An analysis of the aforementioned information is compiled, which is then translated into graphical trade-off maps. The maps illustrate the relevant issues, their importance to each party and trade-off capabilities of each issue. The system takes into account the dynamics of negotiation by representing the relations that exist between issues. Maps are developed by the system to show a negotiator’s preferences and relation strengths between issues. It is from these maps that trade-offs and compromises can be enacted, resulting in changes to the initial values placed on issues. The user is asked if the issues can be resolved in its current form. If this is the case, the system then proceeds to allocate the issue as desired by the parties. Otherwise, the user is asked to decompose an issue chosen by the system as the least

contentious. Essentially the issue on which there is the least disagreement (one party requires it greatly whilst the other party expresses little interest in the issue) is chosen to be the issue first considered. Users are asked to enter sub-issues. As issues are decomposed, they are stored in a decomposition hierarchy, with all links intact This structure has been put in place to recognise there may be sub-issues within issues on which agreement can be attained. It is important to note that the greater the number of issues in dispute, the easier it may be to allocate issues, as the possibility of trade-offs increases. This may seem counter intuitive, but if only one issue needs to be resolved, then suggesting trade-offs is not possible. This process of decomposition continues through the one issue, until the users decide the current level is the lowest decomposition possible. At this point, the system calculates which issue to allocate to which party, then removes this issue from the

each of the party’s respective trade-off maps, and makes appropriate numerical adjustments to remaining issues linked to the issue just allocated. The resulting trade-off maps are displayed to the users, so they can see what trade-offs are made in the allocation of issues. When all issues are allocated at the one level, then decomposition of issues continues, re-commencing from the top level in a sequential manner. Smartsettle (Thiessen and McMahon 2000) assists parties to overcome the challenges of conventional negotiation through a range of analytical tools to clarify interests, identify tradeoffs, recognise party satisfaction and generate optimal solutions). The aim is to better prepare parties for negotiation or to support them during the negotiation process. The algorithms implemented in the system support the process of negotiation by introducing importance values to indicate the degree to which each party desires to be awarded each issue. It is assumed that the importance

value of an issue is directly related to how much the disputant wants the issue to be awarded to her. The system uses this information to form trade-off rules. Systems such as Family Winner are offer far more negotiation support than decision support systems that advise upon BATNAs. 2.4 The Family Law Domain – Negotiation or Justice In section 3.2 we will discuss a general evaluation of negotiation support systems However, at the moment we will confine ourselves to an evaluation of Family Winner. On Tuesday December 3 2003, we met with a number of family solicitors at Victoria Legal Aid (VLA) to evaluate the performance of the Split Up system. Whilst the solicitors were very impressed how Family Winner suggested trade-offs and compromises, they had one major concern – that Family Winner in focusing upon mediation had ignored issues of justice. (Alexander 1992) has illustrated that women tend to be more reluctant than men to continue conflict and are more likely to wave their legal

rights in a mediation session. If their major goal is to be the primary care giver for their children, they may reach a negotiated settlement, which whilst acceptable to them is patently unjust. The wife may Source: http://www.doksinet for example, give the husband the bulk of the property, in return for the wife being the primary care giver of the children. Whilst such an arrangement may meet the goals of both parents, it does not meet the paramount interests of the children, who will be deprived of subsequent financial resources. Family Law is one domain where mediation conflicts with notions of justice. In such domains, the use of negotiation support systems which attempt to equally satisfy both parties, is limited. In the King Solomon judgement, Solomon was able to determine the true mother of the child by assuming no ‘true mother’ would want to divide (and thus kill) her child. But unfortunately many parents are more interested in winning a dispute with their ex-partner,

than in looking after their children’s paramount interests. Suppose two parents both want primary residence of their children, to the exclusion of any other matters or indeed, the interests of the children. Although a compromise, acceptable to both parents, might be to have the children move households every night, no judge would sanction such a course of action, since it would be detrimental to the children. The judge (who is the final arbiter of the paramount interests of the children) can over-ride an agreement negotiated by the parents. Indeed, since legal professionals are aware that judges are unlikely to approve of a settlement where children move house every night, they would caution their clients against even proposing such a solution. However, we have noticed that various bargaining domains are far more suitable than family law, for modeling using integrated game theory and knowledge based systems to advise upon trade-offs 2.5 The Extension of Family Winner to non

Family-Law negotiating domains Given our research on evaluating the Family Winner system, we realise that we need to be careful in choosing negotiation domains that are amenable to the use of decision support systems. In particular we require domains which involve the use of trade-offs and compromises. Since Family Winner has been designed with a view to being utilised in many negotiating domains, there are no domain-specific requirements that prospective cases need to exhibit. In (Bellucci 2003) we discuss how Family Winner has been used in a variety of negotiation domains. Family Law is a less suitable domain for building Negotiation Support Systems than is Enterprise Bargaining. An enterprise bargaining agreement formed our civil law case. It was based on the 2001 Victorian Legal Aid enterprise bargaining agreement. The result of the negotiated settlement was similar to the advice given by Family Winner (Bellucci 2003) discusses two International disputes: an analysis of the Panama

Canal treaty of 1974 and a negotiation held with terrorists. The Panama Canal treaty involved a lengthy series of negotiations between the United States of America and the Panamanian Government. Family Winner’s advice varied considerably from the settlement obtained from negotiations held between the parties. As Family Winner only has the ability to allocate issues to one of the parties, a comparison based upon creative settlements is not possible. Also, there needs to be a better understanding of exactly what issues are in contention, and how these issue definitions can be translated into issues ready for allocation by Family Winner. This example displays one of the drawbacks in using Family Winner on a domain that is quite complicated. The terrorist negotiation example proved that even critical situations can benefit from the use of an automated negotiation system. Another example was a negotiation held between two companies discussing a company merger. The results obtained from

Family Winner in this example demonstrate the effectiveness of point assignments to show the importance value of an issue to a party, coupled with trade-off equations to assist in the allocation of issues valued closely by the respective parties. 3. Future Work and Conclusion 3.1 Building Industry Advisor At Glasgow Caledonian University we are about to commence a project on the Development and testing of a United Kingdom web-based decision support system for use in improving the consistency and predictability of adjudicators’ decisions in building construction disputes. In this project we aim to build a web-based decision support system by Source: http://www.doksinet (a) Combining the records of project partners (The Adjudication Reporting Centre; James R Knowles plc, Construction Contracts Consultants; MacRoberts, Solicitors; and Bishops, Solicitors); electronically publishing these records; creating a standard hub where stakeholders can record adjudication data; and data

mining the records; (b) decision modeling of the domain of building industry dispute resolution by developing a web-based model of legal reasoning in adjudication; (c) comissioning a tool for predicting the course of building dispute adjudications. The project has as objectives: 1. The data mining of adjudication records. The decisions of adjudicators will be examined for coherence The major impact of this will be to replace the largely anecdotal experience of adjudication to a systematic and transparent analysis. Data will be collated into categories that will serve for statistical analysis, much as is currently done by the Adjudication Reporting Centre, but will also be collated from the point of view of the legal issues concerned and how these were dealt with by the adjudicator 2. The development of new web-based, analytical tools to identify the predictability of decisions involving many variables, some of which will depend on legal reasoning, and some on purely statistical

analysis. Accordingly, a decision model will be created based on a legalistic approach (rule based and case based) to adjudication, as derived from over one hundred decisions that have been made by the UK courts. This will enable stakeholders to access the model to anticipate likely outcomes in an adjudication, and for adjudicators to test their decisions. In addition, a model of knowledge discovery will be developed from statistical data, quite apart from the supposed paramouncy of legal reasoning, to identify areas which are likely to affect settlement. 3. Testing the decision model against new adjudications which are being handled by project partners, and the identification of issues where the model aids predictability, and which may be used to facilitate settlement and reduce conflict. The impact of the achieving of these objectives will be to have greater transparency as to the cause of disputes which go to adjudication and the likely result of such disputes. Predictability

should serve to remove part of the expense of projects which is as a result of legal wrangling. The beneficiaries will be all stakeholders in the construction industry, but particularly smaller companies for whom adjudication was designed to help in the first place, thereby helping smaller companies to adopt business improvement measures. There are also teaching and self-help training applications, and the web site would provide a one-stop source of trend data on adjudications. The project will also provide an easily accessed, one-stop decision support system with information on procedure and case law and is capable of being used by adjudicators to test reasoning before publishing a decision. The opportunities it will bring include: (1) Ability to test reasoning; (2) Ability to test likelihood of success; (3) Information and guidance on procedure, thereby reducing the danger of procedural error by adjudicators; (4) Up-todate information and guidance on matters that frequently come to

adjudication; (5) Up-to-date information and guidance on case law; (6) Training tool; (7) Reporter on trends There are national and international markets for such a tool (especially British Commonwealth markets); and the software, the web-based legal modeling concepts, and the overall service offered should have scope for generic reapplications in different jurisdictions. 3.2 Evaluating negotiation support systems Currently we are extending the techniques developed in Family Law to the domain of Labour Law, in particular about contract negotiations. We are also involved in an evaluation of the Family Winner system We use the The Context, Criteria, Contingency (C, C, C) evaluation framework (Hall and Zeleznikow 2001). The CCC framework addresses four areas. • Context: assessing the context of system operation i.e usage • Context: assessing the context of the evaluation itself • Criteria: a hierarchical four quadrant model of evaluation criteria Source: http://www.doksinet

• Contingency: guidelines for the selection of appropriate criteria to satisfy varying evaluation contingencies Initially an evaluator requires to gain a thorough understanding of the context in which the system operates. This usage context considers the parent organization, the application domain, management support, funding, degree of risk exposure, system resourcing and constraints, and the work environment. User issues such as values, motivation, skills, experience and training are also important. An understanding of these issues allows the Evaluator to better appreciate the system context and assists in selecting appropriate criteria to be used in the evaluation. The evaluator also needs to understand the context of the evaluation itself. The evaluation context considers the evaluation feasibility, resources available, constraints applied and the autonomy permitted to the Evaluator. Micro Technology V and V User credibility People Technical Infrastructure Impact Macro

Figure. 4 Evaluation Criteria Model Figure 4 illustrates the Evaluation Criteria model included in the C. C C evaluation framework This model organises potential evaluation criteria into four quadrants bound by two axes: People / Technology and Micro / Macro (small and focussed / large and generalised These criteria have been mapped onto the appropriate quadrant and arranged in a hierarchical fashion. • The Validation and Verification (V and V) quadrant is concerned with the Micro/Technical aspects of technology and the system development process • The Technical Infrastructure quadrant is concerned with the Macro/ Technical infrastructure requirements such as the technical fit of the system with existing systems, resource requirements and availability, portability etc. • The User credibility quadrant is Micro/People oriented and includes User satisfaction, Utility (fitness for purpose, usefulness) and Usability (ease of use). • The Impact quadrant is Macro/People

oriented and is concerned with impact of the system upon its environment, tasks, people, parent organisation and beyond. This model suggests potential evaluation criteria but leaves the evaluation team to make the choice of appropriate criteria, metrics, methods and acceptance level. The evaluation contingency guides their choice Contingency guidelines are still under development and have been drawn from many sources. The long-term goal of the evaluation framework project is to develop guidelines that can be used to frame evaluations of legal KBS under differing contingencies. This paper has considered how negotiation decision support systems can provide useful advice re the reduction of risk in legal disputes. The use of techniques such as argumentation, game theory and heuristics can be of significant assistance. We have provided examples of such systems in the building industry and family law disputes We have also investigated a fundamental conflict between negotiation support

systems and legal decision support systems. When providing negotiation support, we wish to ensure that both parties are equally (dis)satisfied with the proposed solution. In legal domains, considerations of justice can be as or more important than those of user satisfaction with negotiation advice. Source: http://www.doksinet 4. References Alexander, R. 1992 Mediation, violence and the family Alternative Law Journal 17(6): 276-99 Bellucci, E. 2003 Developing Compensation Strategies for the construction of Negotiation Decision Support Systems. PHD thesis, La Trobe University, Bundoora 3086, Victoria, Australia Bellucci, E. and Zeleznikow, J 1997 FamilyNegotiator: an intelligent decision support system for negotiation in Australian Family Law. Proceedings of the Fourth Conference of the International Society for Decision Support Systems, Lausanne, International Society for Decision Support Systems: 359-373. Bellucci, E. and Zeleznikow, J 1998 A comparative study of negotiation

decision support systems Proceedings of the Thirty-First Hawaii International Conference on System Sciences. IEEE Computer Society, Los Alamitos, Cal: 254-262. Bellucci, E. and Zeleznikow, J 2001 Representations for decision making support in negotiation Journal of Decision Support. 10(3-4) Brams, S. J and Kilgour, D M, 2001Fallback Bargaining, Group Decision and Negotiation, 10(4): 287-316 Brams, S. J and Taylor A D, 1996 Fair Division, From cake cutting to dispute resolution Cambridge University Press, Cambridge, UK. Chung, W. W C and Pak, J J F, 1997 Implementing Negotiation Support Systems: Theory and Practice,” presented at Thirtieth Annual Hawaii International Conference on Systems Science - Mini track on Negotiation Support Systems, Hawaii. Clark, E. and Hoyle, A 2002 On-line Dispute Resolution: Present Realities and Future Prospects, 17th Bileta Conference, Amsterdam, <http://www.biletaacuk/02papers/hoylehtml> Douglas, R. and Toulson, D 1999 WIRE Intelligent Quantum

(WIRE IQ) – Tort Evaluation by Precedent instead of ‘Rules’. Proceedings of Twelfth International Conference on Legal Knowledge Based Systems,, GNI, Nijmegen, Netherlands: 127-128. Eidelman, J.A 1993 Software for Negotiations, Law Practice Management, 19(7): 50-55 Eisenberg, M.A, 1976, Private Ordering Through Negotiation: Dispute Settlement and Rulemaking, Harvard Law Review, 89: 637-681. Fayyad, U., Piatetsky-Shapiro, G and Smyth, P 1996 From Data Mining to Knowledge Discovery in Databases AI Magazine, 17(3): 37-54. Fisher, R. and Ury, W 1981 Getting to YES: Negotiating Agreement Without Giving In Boston: Haughton Mifflin Goldring, J., 1976, Australian Law and International Commercial Arbitration Columbia Journal of Transnational Law, 15: 216-252. Hall, M. J J and Zeleznikow, J 2001 Current inadequacies in the evaluation of legal Knowledge-Based Systems Strategies towards a broad-based evaluation model. Proceedings of the Eighth International Conference on Artificial

Intelligence and Law, St. Louis, Missouri: ACM Press: 147-156 Ingleby, R., 1993 Family Law and Society, Butterworths, Sydney Jennings, N.R, Faratin, P, Lomuscio, AR, Parsons, S, Wooldridge, MJ and Sierra, C 2001 Automated Negotiation: Prospects, Methods and Challenges, Group Decision and Negotiation, 10 (2):199-215. Source: http://www.doksinet Katsh, E. and Rifkin, J 2001 Online Dispute Resolution: Resolving Conflicts in Cyberspace Jossey-Bass, San Francisco Ca. Kersten, G.E 1997 Support for Group Decisions and Negotiations, in: J Climaco (Ed) An Overview, in Multiple Criteria Decision Making and Support, (Heidelberg: Springer Verlag). Kolodner, J.L and Simpson, R L, 1989 The Mediator: Analysis of an Early Case-Based Problem Solver, Cognitive Science, 13:507-549. Llewellyn, K. 1962 Jurisprudence University of Chicago Press Matwin, S., Szpakowicz, S, Koperczak, Z, Kersten, G E and Michalowski, G 1989 NEGOPLAN: An Expert System Shell for Negotiation Support, IEEE Expert 4:50-62.

Peterson, M. and Waterman, D A, 1985 Evaluating civil claims: An expert systems approach to evaluating product liability cases in Walter, C. (ed) Computer Power and Legal Reasoning, West Publishing Company: 627659 Ross, H. L, 1980, Settled Out of Court, Aldine Stranieri, A, Zeleznikow, J, Gawler, M, and Lewis, B, 1999. A Hybrid rule- neural approach for the automation of legal reasoning in the discretionary domain of family law in Australia. Artificial Intelligence and Law 7 (2-3): 153183 Stranieri, A., Zeleznikow, J and Turner, H 2000 Data mining in law with association rules Proceedings of IASTED International Conference on Law and Technology(Lawtech2000), ACTA Press, Anaheim, California: 129134. Sycara, K. 1990 Negotiation planning: An AI approach, European Journal of Operations Research, 46:216-234 Sycara, K. 1998 Multiagent Systems, AI Magazine 19(2): 79-92 Thiessen, Ernest M. and McMahon, Joseph P (2000) Beyond Win-Win in Cyberspace Ohio State Journal on Dispute Resolution, Vol.

15, 643 Toulmin, S. 1958 The uses of argument Cambridge: Cambridge University Press Waterman, D. A, Paul, J and Peterson, M, 1986 Expert systems for legal decision making, Expert Systems 3 (4): 212-226. Wilkenfeld, J., Kraus, S, Holley, K M and Harris, MA 1995 GENIE: A decision support system for crisis negotiations, Decision Support Systems, 14:369-391. Williams, G.R, 1983, Legal Negotiation and Settlement, West Publishing Co, St Paul, Minnesota Zeleznikow, J. and Hunter, D 1994 Building Intelligent Legal Information Systems: Knowledge Representation and Reasoning in Law, Kluwer Computer/Law Series, 13. Zeleznikow, J., Stranieri, A and Gawler, M 1996 Split-Up: A legal expert system which determines property division upon divorce. Artificial Intelligence and Law 3:267-275 Zeleznikow, J., Meersman, R, Hunter, D and van Helvoort, E 1995 Computer tools for aiding legal negotiation ACIS95 Sixth Australasian Conference on Information Systems, Curtin University of Technology, Perth, Western

Australia: 231-251