The Worlds of Negotiation

The Worlds of Negotiation Mareike Schoop Information Systems, University of Hohenheim, Germany [email protected] Abstract Traditionally, electro...
Author: Jane Ford
13 downloads 0 Views 238KB Size
The Worlds of Negotiation Mareike Schoop Information Systems, University of Hohenheim, Germany [email protected] Abstract Traditionally, electronic negotiations have focused on automation of processes. Recently, it was argued that a communication perspective on electronic negotiations needs to be supported in order to achieve the aim of enabling even complex negotiations electronically. We will argue that there is a need to combine these two classes of approach with another class, namely that of decision support and that a holistic integration is required to achieve the goal of electronic negotiation support. Our arguments will be based on Habermas' view of different worlds and will be illustrated using the Negoisst system.

1

Introduction

Electronic negotiations are an essential part of a business transaction which consists of three phases, namely searching for new information, negotiating the contractual details, and fulfilling the business contract (Schmid and Lindemann, 1998; Schoop, Köller, et al., 2001). When looking at the field of negotiation support, three different classes of approach can be distinguished. Firstly, there are the automation-oriented approaches aiming at automating the negotiation process in order to find an economic optimum. Secondly, there are the communication-oriented approaches supporting the communicative processes involved in any negotiation. Thirdly, The copyright of this paper belongs to the paper’s authors. Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage. Proceedings of the 9th International Working Conference on the Language-Action Perspective on Communication Modelling (LAP 2004) Rutgers University, The State University of New Jersey, New Brunswick, NJ, USA, June 2-3, 2004 (M. Aakhus, M. Lind, eds.) www.scils.rutgers.edu/lap04/lap04.htm

179

M Schoop there are the document-oriented approaches enabling document exchange and document management. While the automated approaches are the dominant class in the field of electronic negotiations, a communication perspective has recently been emphasised (Heng and Moor, 2003; Weigand, Schoop, et al., 2003) in order to account for the fact that there are complex negotiations that cannot be automated but that require effective communication support (Schoop, Jertila and List, 2003).

2

Classifications of Electronic Negotiations

A literature search shows that there are various proposed classifications of electronic negotiations. The most relevant classification approaches will now be introduced and later revisited. Negotiations can be distributive or integrative (Kersten, 2001; Lewicki, Saunders and Minton, 1999; Walton and McKersie, 1965; Weigand, Schoop, et al., 2003). Distributive negotiations aim at maximising the outcome for each partner. Obviously, achieving a gain for one negotiator can sometimes only be achieved at the expense of the other partner. Integrative negotiations aim to achieve a deal that reflects a good outcome for both parties. Integrative approaches are also called “win-win approaches” as opposed to the “win-lose approaches” represented by distributive models. Research approaches to electronic negotiations can be classified according to negotiation theory into normative, prescriptive, and descriptive approaches (Bell, Raiffa and Tversky, 1991; Kersten, 2002). Normative negotiations models are concerned with the design of models of rational negotiators and their procedures of interaction. Such models aim at economic rationality and have been used in gametheoretic approaches where the outcome is an efficient and stable compromise (Lomuscio, Wooldridge and Jennings, 2001; Munier, 1993; Peters, 2000). Prescriptive negotiation models are concerned with helping negotiators achieve good outcomes by prescribing procedures and user behaviour. These approaches have been used in decision and negotiation analysis and management sciences (Bui, 1994; Jarke, Jelassi and Shakun, 1987; Kersten, 1988; Teich, 1996). Descriptive negotiation models are concerned with understanding user behaviour, in particular problem situations, negotiation patterns, reasons for certain decisions, cultural differences in negotiation behaviour, and with the implications for designing processes and systems based on these experiences(Bazerman, Curhan, et al., 2000; Gulliver, 1979; Kersten, Köszegi and Vetschera, 2002; Neale and Bazerman, 1991). These approaches have been predominantly used in behavioural science, psychology, and sociology. Kalakota (1997) discusses different perspectives on electronic commerce, e.g. organisational perspective, process perspective, technological perspective. There are also different viewpoints concerning electronic business negotiations. The two most relevant in the present context are the computer science viewpoint and the

180

The Language-Action Perspective on Communication Modelling 2004

The Worlds of Negotiation economic viewpoint. Looking at electronic negotiations from a computer science point of view, the main aim is to develop IT support for the negotiations. On the one hand, the support concerns the data management. Since the negotiation phase is characterised by a high level of dynamics and interaction, there is a large amount of data that is created and that needs to be stored, accessed, retrieved, and processed efficiently. On the other hand, the process itself can be supported by information technology. For example, workflow systems can control the order of the process and enforce a certain negotiation protocol; software agents can take over parts of the negotiation process; reasoning about the results of the negotiation process can be enabled; an architecture for a negotiation system can be developed. Therefore, the computer science view emphasises prescriptive approaches. In the case of workflow systems, a certain structure of the process is prescribed and enforced. In the case of data management, the prescriptive approach is taken in a more instrumental way. Looking at negotiations from an economic view, the main aim is to achieve economic efficiency. For example, automating the negotiation process results in reduced transaction costs; computing the optimal outcome helps to achieve better deals for the participants. Thus, normative approaches are emphasised. Approaches in both computer science and economics need a thorough understanding of the processes to be supported or automated, of the negotiation patterns to be supported or replaced, of the user requirements in terms of technical and economic goals, and of the rationality guiding the negotiation processes. Thus, some descriptive research is necessary as a starting point. The negotiation process itself can have different orientations(Weigand, Schoop, et al., 2003). The first dimension is concerned with the motivation for agreement. If a negotiator acts on the basis of obligations and social commitments resulting from negotiation steps, we speak of norm-based negotiations. If a negotiator acts on the basis of expressed goals and objectives, we speak of goal-based negotiations. Norm-based negotiations are the most common type. For example, requests, offers, quotes, and counter-offers are examples of speech acts that presuppose some kind of norm shared by the negotiators. A quote can be interpreted as an authorisation for the customer to order as specified in the quote. Goal-based negotiations are less known in the Western society and are closer to the Oriental cyclic style. The negotiator is not motivated by norms but by goals (s)he tries to achieve. A deal is struck when the goals of the negotiators are met sufficiently. Trust plays a more important role since there are less standard processes to follow. Thus, it has been argued that such type of negotiations occur in closed communities or virtual organisation where the trust issue has already been dealt with. Electronic negotiations can occur in different forms which can depend on the type of application used and the type of electronic commerce. In business-toconsumer e-commerce, negotiations are less known whereas negotiations play an important role in business-to-business electronic commerce. Portals aim at

The Language-Action Perspective on Communication Modelling 2004

181

M Schoop providing information regarding one topic. The search for such information is of prime importance and no other transaction phase is supported. Thus, negotiations do not exist in portals. Shops do usually not offer negotiation facilities since the prime functionality is catalogue shopping which means fixed prices for all products. One might imagine the possibility to bargain in electronic types of shops as well but this application area is not of prime importance for negotiations. The goals of supporting negotiations through information technology can be manifold. For example, it is possible to conduct checks during the negotiation (Schoop, Jertila and List, 2002) or to offer decision support (Kersten and Noronha, 1999; Kersten and Lo, 2001; Kersten and Lo, 2002). Automated models usually aim at reducing transaction costs (Peters, 2000; Rebstock, 2001) and/or at finding an optimal deal (Bichler, 2001; Maes, Guttman and Moukas, 1999; Sandholm, 1999). Other models provide argumentation support for human or software agents (Karacapilidis and Papadias, 1998) or aim to enhance the communicative quality of exchanges by supporting and disambiguating negotiation utterances (Schoop, 2002; Schoop, Jertila and List, 2003; Weigand, Schoop, et al., 2003). In general, information technology enables asynchronous negotiations (in contrast to face-toface interactions), can help to manage negotiation data efficiently, and can prescribe a particular negotiation process or protocol to follow.

3

Communicative Worlds

In his Theory of Communicative Action (Habermas, 1985), Habermas argues that pragmatic aspects of language must be considered in addition to syntax and semantics. Habermas takes Searle’s Theory of Speech Acts (Searle, 1969) as a starting point for his own theory reflecting his concerns with pragmatic aspects of language. Habermas proposes a set of rules that govern communicative competence. Like Searle he argues that each speech act consists of an illocutionary force and a propositional content. The illocutionary force establishes the mode of communication between speaker and hearer and thus the pragmatic situation of the contents of the utterance. The propositional content establishes the relation between the utterance and the outside world. Habermas distinguishes between three types of worlds or realities. The external reality or objective world consists of the world of objects and events. It is what is usually perceived as the reality. The relevant validity claim is that of truth. The internal reality or subjective world is the speaker’s own world of experiences. The relevant claim is that of truthfulness. The social world or normative world is the world of norms and values and that of relevant commitment. The relevant validity claim is that of appropriateness. In an utterance, a speaker is seen as communicating about the internal or external world in the context of a given social/normative world. Whereas the subjective world concerns one being only,

182

The Language-Action Perspective on Communication Modelling 2004

The Worlds of Negotiation the objective world and the social/normative world are shared worlds. Nevertheless, the perspective on reality and on a normative context can be very different so that communication is necessary for the communication partners to find a common background. The semiotic triangle is a well-known concept in semiotics and can be linked to Habermas’ notions of communicative worlds. The semiotic triangle distinguishes between objects, concepts, and terms, see figure 1. An object is any part of the objective world (both that part that really exists and that part that can be imagined). During our socialisation, we have learned to classify elements of the objective world into abstract classes. Thus, we derive concepts that are units of thoughts consisting of a set of objects with similar attributes. A concept is part of our internal reality and only exist in our thoughts. In order to communicate, we need a designator or term that designates the concept and thus denotes the object. Such a term is the representation of a concept by language or other means. The term is related to the social/normative world as terminology is part of our world of linguistic norms and standards. The process of communication thus concerns all three types of worlds. Concept

designates

relates to

Term

Object denotes

Figure 1. The semiotic triangle

4

The Worlds of Electronic Negotiation

As argued in the previous section, communication is a process concerning the objective world as the basis of communicative content through the observation of elements or actions, the subjective world as the basis of communication through the ability of recognising relations between observed objects to form classes and relations, and the normative world as the basis of the illocutionary force representing the communicative content. Since negotiations are a type of communication, a negotiation process will be concerned with all three worlds as well. Consequently, electronic negotiations require support that takes into account the three worlds and their interrelations. 4.1

Challenges

The Language-Action Perspective on Communication Modelling 2004

183

M Schoop Our aim is to enable efficient and effective electronic negotiations. Firstly, we will briefly review the limitations of the three classes of negotiation support, namely automated approaches, communication-oriented approaches, and documentoriented approaches. Secondly, we will discuss requirements for holistic negotiation support w.r.t. prescriptive, descriptive, and normative models. Thirdly, we will critically review the meaning of integrative or distributive strategies for practical negotiation support. While automated approaches have been popular in recent times (Bichler, 2000; Sandholm, 1999), they have certain limitations that show well why they cannot deal with complex negotiations (Kersten, Noronha and Teich, 2000; Schoop, Jertila and List, 2003). The main limitation is that the automated approaches enable trading only of standard products that can be described in detail and which are mostly invaluable goods. There is no dynamic product specification or communicative enrichment process during the negotiation. The product characteristics are fixed at the beginning of a negotiation process to enable comparability of requests and offers. Automated approaches follow strict protocols that do not allow for flexible interactions tailored to specific requirements of particular negotiation context. Reducing a negotiation to a strictly rational process means that certain processes requiring complex communication (e.g. when discussing offers or justifying a particular counter-offer) cannot be represented. As a counter-movement to automated approaches, both communication-oriented and document-oriented approaches have been proposed. They deal with the support of negotiators rather than the automation of processes. Nevertheless, they also have a number of serious disadvantages. The main limitation of communication-oriented approaches is that the message content is unstructured. Therefore, there is no possibility of structured access to the content of a written utterance representing an offer, a request, a counter-offer, a quote etc. Furthermore, there is no support of business documents. The main limitation of document-oriented approaches is that important information about the negotiation process is lost. Only the outcome of a negotiation process is stored (namely the contract documents) but the history behind it is lost. Therefore, the first challenge is that there must be an integrative approach to electronic negotiations containing elements from all three classes. Communication is an important part of any negotiation process and thus needs to be supported. Contract documents are the second important part of a negotiation and should be managed. The automation of certain parts of the processes might be beneficial for the negotiators. Furthermore, the automation approaches can deal very well with multi-attributive quantitative information that is relevant for a negotiation approach. In particular, all three worlds according to Habermas’ theory need to be represented. Any study of communicative behaviour (and thus of negotiations as a specific type of communication) starts with a phase of requirements analysis and analysis

184

The Language-Action Perspective on Communication Modelling 2004

The Worlds of Negotiation of current practices. Therefore, there will always be a descriptive element as the basis of future innovations. The more prescriptive an approach is, the easier it is to support it electronically but the less flexible the process is. Normative approaches aim at modelling rational negotiators interacting in a normative standardised way. While there are a number of standards for electronic negotiations (such as EDIFACT of eBXML), they can only be applied to well-structured scenarios in which the negotiation elements are fixed and there is no need for discussions or complex communicative steps. Therefore, the second challenge is to support the necessary descriptive, prescriptive, and normative elements of a negotiation while retaining sufficient flexibility to allow for complex electronic negotiations that require flexible communicative interactions. It has been argued that integrative negotiations are to preferred to distributive negotiations since they leave negotiators happier with the outcome and thus motivate them to negotiate again with the same partner and they achieve good deals for both partners (Kersten, 2001). In Habermas’ terminology, integrative negotiations are a type of communicative action where the motivation is a rationally grounded consensus between the negotiation partners. Such negotiations have been classified as “win-win” negotiations. Communicative actions are oriented towards an agreement between an author of an utterance and its recipient. In contrast, distributive negotiations are a type of strategic actions which follow rational rules and try to influence a rational opponent in order to achieve personal success. Such negotiations have been classified as “win-lose” negotiations. Strategic actions are oriented towards success. Here, “success” means that the author succeeds in getting the hearer to behave in a way that suits the author’s own goals. Following Habermas’ arguments, distributive negotiations would be defective speech acts since they do not follow the ideal of communicative action. Such argument has to be taken with some caution. While integrative negotiations can certainly achieve good outcomes for both partners (e.g. pareto-optimal outcomes), there exist negotiation strategies that aim at achieving a personal maximum and that are applied successfully in certain situations, e.g. when there are one-off interactions where there is no need to make a partner happy so that future interactions will take place in the same constellation. Furthermore, when there are scarce resources or there is scarce demand for a good, then the seller or buyer respectively are in a strong position and will usually use a distributed strategy since it is obvious that such a strategy will lead to better results in these cases. Therefore, the third challenge is to support both integrative and distributive negotiations. This challenge is related to the second challenge. While we could use a normative approach that prevents a distributive negotiation to take place, this would exclude some forms of negotiation that take place in the traditional form

The Language-Action Perspective on Communication Modelling 2004

185

M Schoop and that should be supported electronically. Therefore, the aim of flexibility applies here as well. To summarise, the aim is to enable a holistic support of electronic negotiations: • enabling communicating about the subjective and objective world based using terms related to professional or societal norms • integrating communication support and document management with a specification of contents containing multi-attributive and also quantitative information • based on a descriptive study of communicative behaviour in traditional and existing electronic negotiations and requirements for effective and efficient electronic negotiation support • using a prescriptive approach in an instrumental manner rather than strictly enforcing certain behaviour • enabling the flexible use of norms and standards with the possibility of refining them during the negotiation process. The following section will introduce our approach that meets these challenges and aims. 4.2

Solutions

In this section, we will introduce our negotiation support system Negoisst that enables efficient and effective support of complex electronic negotiations between human negotiators. The different features of Negoisst will be presented before we will discuss how the challenges presented in the previous section have been met. 4.2.1 Message exchange The message is the most important unit in Negoisst; it is the medium for the negotiators to conduct the actual negotiation. In direct negotiations, the other partner can be observed in terms of mimics, gestures, tone of voice, actions etc. which helps to interpret an utterance and to understand a speaker’s intentions. In electronic interactions, such a direct interpersonal contact is removed. Thus, there must be some different mechanisms to make the speaker’s intention explicit and thus to prevent ambiguity of this part. These ideas are influenced by media richness theory. In order to enrich the medium of written interactions, there are two mechanisms. Firstly, the message type is specified for each message by the sender. There are seven message types based on a refined version of Searle’s classification of speech acts: offer, request, counteroffer, accept, reject, question, clarification. The message type determines the negotiation process and is further used to derive obligations resulting from agreements.

186

The Language-Action Perspective on Communication Modelling 2004

The Worlds of Negotiation Secondly, there is a semantic enrichment process of the message content. The content of a negotiation step is specified in natural language as the most intuitive form of interaction. Since natural language can be ambiguous and since such ambiguity can become a serious problem in written interactions where mimics and gestures cannot help to interpret an utterance (Schoop, Jertila and List, 2003), we created a mechanism that would retain the flexibility of natural language messages while providing the means for unambiguous utterances. This mechanism is the semantic enrichment. The negotiator can highlight parts of the text and associate them with a semantics by choosing predefined attributes. This leads to a reduction of unintended ambiguity. Figure 2 shows such an enrichment process. The negotiator writes the message as shown in the left window and wants to specify that 31 March 2004 is the reply date (and not, for example, the delivery deadline or the date of payment). (S)he then highlights the extract “31 March 2004” and choose the correct ontology elements representing a contract point as shown in the middle window where the date of reply is then shown. This information is used to automatically create the contract version as will be discussed in section 4.2.3. The messages themselves are stored in a relational database.

Figure 2: Semantic enrichment process

The Language-Action Perspective on Communication Modelling 2004

187

M Schoop

4.2.2 Ontology Negotiations The list of contract points is predefined. However, there will always be situations in which the list is not sufficient to express the correct meaning. Therefore, there must be a possibility to negotiate about the background of the negotiation, namely the underlying ontology. The process of ontology refinement is a negotiation in itself where the user can propose a change and the receiver can only accept or reject it. The message content field can contain all communicative steps that are required to agree on a revised ontology. Once the negotiators have agreed on the changes to the ontology, the result will be a consistently adapted ontology that can be the basis for future exchanges or for the particular negotiation only. The users remain in control over the negotiation process as well as over their ontological basis for the negotiation. Figure 3 shows an extract of an ontology negotiation. One negotiator wants to negotiate about the material of the goods in question (in this case skirts) which is not part of the existing ontology. She then proposes a change to the ontology and creates “material” as a new attribute of skirt. If the negotiation partner agrees, the ontology will be adapted accordingly and the material of a skirt can now be used in the negotiation. For example, “silk” could then be specified in the contract negotiation and would thus represent an instance of the material of a skirt.

Figure 3: Ontology negotiation

188

The Language-Action Perspective on Communication Modelling 2004

The Worlds of Negotiation The idea of ontology negotiations is closely related to general Semantic Web ideas (Schoop, Becks, et al., 2002). However, while the semantic web approaches mainly concentrate on the search side, our idea is to use the ontological context of the semantic search and re-use the same context for the negotiation. Ontologies are modeled in OWL as the current standard for the Semantic Web. Figure 4 shows the use of ontologies in Negoisst.

Figure 4: Ontologies in Negoisst 4.2.3 Document exchange The goal of a negotiation is to develop an agreement documented in a business contract. Each message leads to a new contract version which means that the business documents are created and managed in parallel to the message exchange. The system automatically extracts the message content, in particular the selected attributes and their values, and includes them in the current contract version. Meta data such as the identity of the negotiation partners and their duties as e.g. supplier and customer, the date of creation, the status (i.e. “contract version” or “contract”) and the negotiation theme are also recorded automatically. Negoisst further creates links between each contractual item and the related message in which that item was defined or modified, thereby ensuring traceability of the history after each negotiation step. Figure 5 shows the contract version resulting from the interaction step that was shown in figure 2. All relevant contract points are shown both with their value and their meaning.

The Language-Action Perspective on Communication Modelling 2004

189

M Schoop

Figure 5: Contract version 4.2.4 Decision support Using the semi-structured semantic enrichment of each message, formal decision support mechanisms can be applied in real-time while messages are composed, thereby building up a valuable history of evaluated negotiation steps. However in complex negotiations, classical decision support assumptions do not hold: often negotiators focus on a single attribute first and then iterate through the contract while fully specified alternatives are required for most decision support mechanisms. During this interaction new issues arise regularly and preferences dynamically evolve because of new information that is acquired in the communication process.

190

The Language-Action Perspective on Communication Modelling 2004

The Worlds of Negotiation Therefore, we implemented decision support and in particular preference elicitation in a flexible way in Negoisst as shown in figure 6 at the bottom. The decision support is also transparent to the negotiators. A Hybrid Conjoint approach is employed to elicit individual preferences which are then used to give a rating of each request, offer or counteroffer made. If only parts of a contract are under discussion, the system gives worst- and best-case utility estimations which can be interpreted as indicators for the flexibility left in the negotiation. Figure 5 shows the rating function for a particular message. The value is a range since there are some as yet un-specified attributes.

Figure 6: Decision support 4.3

Discussion

Negoisst is a negotiation support system implementing theoretical elements from our research on communication theory, semantic web, decision support, and document management. It enables a holistic communication process. Negotiators are able to express their subjective communicative elements based on the objective world in an

The Language-Action Perspective on Communication Modelling 2004

191

M Schoop intuitive way by using natural language. Furthermore, the norms underlying such exchanges can also be the subject of negotiations. Therefore, the communicative norms themselves can be defined, refined, modified, proposed, accepted, or rejected by the negotiation partners which enables negotiations tailored to particular needs. On the other hand, the ontologies as the basis of any negotiation represent general norms. The last point is important to mention since the aim of negotiation support must not be to enable interactions that are specific to particular partners and thus require new standards for interactions with other partners. This has been the problem of many standards such as EDIFACT. The quantitative world is also important in negotiations and is integrated with document management and communication support in Negoisst. Therefore, negotiators can choose appropriate support for their present needs. An integrative approach has thus been implemented in Negoisst and fulfils the first challenge as presented in section 4.1. The research basis of Negoisst consists of various descriptive studies. For example, requirements have been analysed in areas such as cooperative building teams, the fashion industry, or management-union negotiations. The prescriptive part of Negoisst is done in an instrumental rather than a strictly enforcing manner, thereby combining communication, document, and process support with flexibility in the individual processes. Thereby, the second challenge is met. There is no distinction between integrative and distributive negotiations. We believe that the negotiators should choose themselves which strategy to apply. Thus, we explicitly support integrative as well as distributive negotiations and thereby meet the third challenge.

5

Empirical Evidence

Negoisst was evaluated in various settings. We have conducted extensive validations in the construction industry where Negoisst was used to conduct negotiations between architects and other trades in cooperative construction teams. Further details are discussed in (Schoop, 2002). Furthermore, Negoisst is currently used in the fashion industry in the context of ontology-based electronic negotiations (Schoop, Becks, et al., 2002; Schoop, Jertila and List, 2003). Negoisst has also taken part in the international negotiation tournaments where different negotiation support systems compete. Finally, an extensive experiment was conducted with a voluntary group of students during a thematically related lecture in January 2004. The participants were undergraduate students of Information Systems and had little experience with formal negotiations, either offline or via an electronic medium. No information was disclosed in advance and all groups were trained during a 30 minute presentation of the system and its features. A pre-defined case was the subject of the negotiation. The results showed general acceptance of different features. In this

192

The Language-Action Perspective on Communication Modelling 2004

The Worlds of Negotiation study, the decision support feature was tested as well as the semantic enrichment process. We will elaborate on the findings in a further publication but we would like to mention some results. The first proposition that was tested is that decision support functionality in NSSs positively contributes to the user expectancy of effort and performance. The second proposition which represents the main focus of Negoisst is that communication structuring in NSSs positively contributes to the user expectancy of effort and performance. Both of the propositions cannot be rejected. For example, it was shown that significant differences existed between the group without decision support (DSS) functionality in the system and the control group which was provided with that functionality. The provision of DSS functionality thus offers additional value to the test persons in terms of expected usefulness and productivity (performance and effort). Consistent with this perception, negotiators with decision support finished their negotiations with fewer messages, which is an indication of productivity. Based on these findings, we are currently working on an efficiency measure for electronic negotiations integrating quantitative and qualitative (especially communicative) features.

6

Conclusion

In this paper, we discussed the world of negotiation based on Habermas’ notion of objective, subjective, and normative world. These world views apply to negotiation as well since negotiation is a specific form of communication. In order to enable effective electronic negotiations, all three worlds need to be considered. We discussed the challenges that complex negotiations pose on any negotiation support system. Then, the negotiation support system Negoisst was presented which is strongly rooted in communication theories and LAP in particular but which is also influenced by other world views, namely decision support, semantic web, and document management. Finally, we dicussed how the challenges are met in Negoisst. Further work consists of empirical research which is largely missing in the area of electronic negotiations. The international negotiation tournaments that compare different negotiation systems are one step towards more empirical data. Other negotiation models such as electronic auctions or negotiation agents will be analysed next. Our overall goal is to integrate all classes of negotiation support into one holistic system that can provide effective negotiation support tailored to the context in question but general enough to be applicable to similar contexts.

The Language-Action Perspective on Communication Modelling 2004

193

M Schoop

References M. H. Bazerman, J. R. Curhan, D. A. More and K. J. Valley (2000). Negotiation. Annual Review of Psychology, 51. 279-314. D. Bell, J. Raiffa and A. Tversky (eds.) (1991). Decision Making: Descriptive, Normative, and Prescriptive Interactions. Cambridge University Press, Cambridge. M. Bichler (2000). A Roadmap to Auction-based Negotiation Protocols for Electronic Commerce. In 33rd Hawaii International Conference on Systems Sciences, (HICSS-33), Maui, Hawaii, IEEE Computer Society Press. M. Bichler (2001). BidTaker: An Application of Multi-Attribute Auction Markets in Tourism. In H. Buhl, A. Huther and B. Reitwiesner eds. Information Age Economy, 5. Internationale Tagung Wirtschaftsinformatik, WI 2001, Augsburg, Germany, Physica-Verlag, 533-546. T. Bui (1994). Designing Multiple Criteria Negotiation Support Systems: Frameworks, Issues and Implementation. In G. H. Tzeng, H. F. Wang, U. P. Wen and P. L. Yu eds. Multiple Criteria Decision Making - Proceedings of the Tenth International Conference: Expand and Enrich the Domains of Thinking and Application, Springer Verlag, 321-330. P. H. Gulliver (1979). Disputes and Negotiations: A Cross-Cultural Perspective. Academic Press, New York. J. Habermas (1985). The Theory of Communicative Action. Beacon Press. M. S. H. Heng and A. D. Moor (2003). From Habermas’s communicative theory to practice on the internet. Information Systems Journal, 13. 331-352. M. Jarke, M. T. Jelassi and M. F. Shakun (1987). MEDIATOR: Toward a Negotiation Support System. European Journal of Operational Research, 31 (3). 314-334. R. Kalakota and A. B. Whinston (eds.) (1997). Readings in Electronic Commerce. Addison Wesley. N. Karacapilidis and D. Papadias (1998). A Group Decision and Negotiation Support System for Argumentation Based Reasoning. In G. Antoniou, A. K. Ghose and M. Truszczynski eds. Learning and Reasoning with Complex Representations, LNAI 1359, Springer Verlag, 188-205. G. E. Kersten (1988). A Procedure for Negotiating Efficient and Non-Efficient Compromises. Decision Support Systems, 4. 167-177.

194

The Language-Action Perspective on Communication Modelling 2004

The Worlds of Negotiation G. E. Kersten and S. J. Noronha (1999). WWW-based Negotiation Support: Design, Implementation, and Use. Decision Support Systems, 25. 135-154. G. E. Kersten, S. J. Noronha and J. Teich (2000). Are All E-Commerce Negotiations Auctions? In Fourth International Conference on the Design of Cooperative Systems, COOP 2000, Sophia-Antipolis, France. G. E. Kersten (2001). Modeling Distributive and Integrative Negotiations. Review and Revised Characterization. Group Decision and Negotiation, 10 (6). 493-514. G. E. Kersten and G. Lo (2001). Negotiation Support Systems and Software Agents in E-Business Negotiations. In C. Lee ed. First International Conference on Electronic Business, Hong Kong. G. E. Kersten (2002). The Science and Engineering of E-negotiation: Review of the Emerging Field. InterNeg Reports INR04/02, Montréal, Canada. G. E. Kersten, S. Köszegi and R. Vetschera (2002). The Effects of Culture in Anonymous Negotiations. Experiments in Four Countries. In J. R.H. Sprague ed. 35th Hawaii International Conference on Systems Sciences, Hawaii, IEEE Computer Society Press, Los Alamitos. G. E. Kersten and G. Lo (2002). Aspire: Integration of Negotiation Support System and Software Agents for E-Business Negotiation. Quarterly Journal of Electronic Commerce. submitted. R. J. Lewicki, D. M. Saunders and J. W. Minton (1999). Negotiation. Third Edition. Irwin McGraw-Hill, Boston. A. R. Lomuscio, M. Wooldridge and N. R. Jennings (2001). A Classification Scheme for Negotiation in Electronic Commerce. In F. Dignum and C. Sierra eds. Agent-Mediated Electronic Commerce - The European AgentLink Perspective, LNAI 1991, Springer Verlag, 19-33. P. Maes, R. H. Guttman and A. G. Moukas (1999). Agents that Buy and Sell. Communication of the ACM, 42 (3). 81-91. B. R. Munier (1993). Are Game-theoretic Concepts Suitable Negotiation Support Tools - From Nash Equilibrium Refinements Toward a Cognitive Concept of Rationality. Theory and Decision, 34 (3). 235 - 253. M. A. Neale and M. H. Bazerman (1991). Cognition and Rationality in Negotiation. Free Press, New York. R. Peters (2000). Elektronische Märkte und automatisierte Verhandlungen. Wirtschaftsinformatik, 42 (5). 413-421.

The Language-Action Perspective on Communication Modelling 2004

195

M Schoop M. Rebstock (2001). Elektronische Unterstützung und Automatisierung von Verhandlungen. Wirtschaftsinformatik, 43 (6). 609-617. T. Sandholm (1999). Automated Negotiation: The Best for All Concerned. Communications of the ACM, 42 (3). 84-85. B. Schmid and M. Lindemann (1998). Elements of a Reference Model for Electronic Markets. In 31st Hawaii International Conference on System Sciences, HICSS 98, Hawaii, IEEE Computer Society Press, 193-200. M. Schoop, J. Köller, T. List and C. Quix (2001). A Three-Phase Model of Electronic Marketplaces for Software Components in Chemical Engineering. In B. Schmid, K. Stanoevska-Slabeva and V. Tschammer eds. First IFIP Conference on E-Commerce, E-Government, E-Business, I3E 01, Zurich, Switzerland, Kluwer Academic Publishers, 507-522. M. Schoop (2002). Electronic Markets for Architects - The Architecture of Electronic Markets. Information Systems Frontiers, 4 (3). 285-302. M. Schoop, A. Becks, C. Quix, T. Burwick, C.Engels and M. Jarke (2002). Enhancing Decision and Negotiation Support in Enterprise Networks Through Semantic Web Technologies. In R. Tolksdorf and R. Eckstein eds. XML Technologien für das Semantic Web - XSW 2002, Berlin, Lecture Notes in Informatics, 161-167. M. Schoop, A. Jertila and T. List (2002). Negoisst - A Negotiation Support System for Business-to-Business Electronic Commerce. Data Knowledge and Engineering, accepted. M. Schoop, A. Jertila and T. List (2003). Negoisst - A Negotiation Support System for Business-to-Business Electronic Commerce. Data Knowledge and Engineering, 47 (3). 371-401. J. R. Searle (1969). Speech Acts - An Essay in the Philosophy of Language. Cambridge University Press. J. E. Teich (1996). Identifying Pareto-optimal Settlements for 2-party ResourceAllocation Negotiations. European Journal of Operational Research, 93 (3). 536 - 549. R. E. Walton and R. B. McKersie (1965). A Behavioral Theory of Labor Negotiations. McGraw-Hill, New York. H. Weigand, M. Schoop, A. D. Moor and F. Dignum (2003). B2B Negotiation Support: the Need for a Communication Perspective. Group Decision and Negotiation, 12 (1). 3-29.

196

The Language-Action Perspective on Communication Modelling 2004