System for simulating a hearing in a procedure, especially a court hearing

Abstract

A computer-implemented system for simulating a negotiation in a procedure, in particular a court hearing, comprising a data module (1) for storing relevant case data; an analysis module (3) configured to assign the data retrieved by the data module (1) to agents of the negotiation in order to generate a respective agent avatar and to establish a negotiation simulation model based on the retrieved data; a simulation module (4) that simulates the negotiation using the negotiation simulation model and the agent avatars; and a user interface (5) via which a user to be trained interacts with the system as an agent of the negotiation.

Description

The present invention is in the field of negotiations with one or more parties. In particular, the present invention contemplates government negotiations or pre-office negotiations, such as the European patent office or other patent offices or other public premises. However, the invention is not limited thereto and can be applied to any negotiation with multiple participating agents.An agent in this sense is one or more persons or an instance that actively participates in a negotiation and influences the course of the negotiation by their actions or decisions. These can be real persons such as parties, their representatives, vehicles or directores. An agent is thus an acting subject within the method that acts on the basis of information, makes decisions, or responds to arguments and counter arguments, thereby contributing to the dynamics of the negotiation.In particular, jurisdiction negotiations, such as with multiple parties, assets, and a jurisdiction board, populated with multiple jurisdictions, become increasingly complex and often require a long preparation time for the party representatives. In order to prepare the negotiation, not only must all fonts and relevant information about the case be analyzed, but a defense or attack strategy must also be worked up. Moreover, careful preparation for negotiation can never fully depict the atmosphere during negotiation and interaction with the appropriate agents, such as director and other process participants.Very important methods are sometimes concerned with simulating such negotiation with real persons in a room. On the one hand, this is cost-intensive, but on the other hand, there are also problems with termination over and over again, since the corresponding persons with whom it is desired to simulate the negotiation are prevented.Proceeding from this and in view of the increasingly varied possibilities of artificial intelligence, the present inventors have set themselves the aim of specifying a computer-implemented system with which such a negotiation can be simulated realistically. This system accesses a large database in particular and has, for example, a AI-Bot module which is able to process the information and train it with a corresponding user or an agent who participates in the negotiation.A Kl-Bot module in the context of the present invention denotes a computer-implemented component which is based on artificial intelligence and is used for simulating a negotiation. This module processes a large database of relevant information, analyzes fonts and strategic aspects, and interacts with users or other agents within the negotiation to simulate the negotiation in advance. It can be used to realistically simulate different negotiation scenarios by recognizing argumentation patterns, simulating responses, and simulating the interaction with process participants, such as director or party. The aim of the AI-BOT module is to provide a training and preparation environment for participants in negotiation that is as realistic as possible, in order to optimize the effectiveness of the negotiation management.To solve the problem, a computer system is provided having the features of claim 1.The present system is used to simulate a negotiation in a method. This may be, for example, a jurisdiction negotiation, but may also be any other negotiation with multiple participating agents, including one-way methods, such as login methods prior to a national or regional patent office or other public offices.The system comprises, for example, a data module, an analysis module, a simulation module and a user interface. Relevant case data are stored in the data module, for example. Via the user interface, a user can specifically train his negotiation capabilities by interacting with the system as a corresponding agent. The term agent is to be understood here in a broad sense and derives from the Latin age, which means "act" or "do something". As described in the introduction to the specification, an agent is thus an acting, in particular real, physical subject within the method which acts on the basis of information, makes decisions or responds to arguments and counter arguments, thereby contributing to the dynamics of the negotiation.The analysis module retrieves the information stored in the data module and assigns it to the respective agents relevant to the specific negotiation. On the basis of this, a corresponding digital avatar is generated for each agent-a virtual representation of the respective person within the simulated negotiation.An agent avatar, in the sense of the above text, denotes a digital representation of an agent within the simulated negotiation. It serves to map a real person, such as a director, a party or a representative of a party, in software and to take over their role in the negotiation. The agent avatar is generated on the basis of the information stored in the data module and processed by the analysis module and reflects the features and behaviors of the respective agent specific to the negotiation.The AI model on which the system is based or the AI module can automatically generate the number of agents and the respective roles, for example, on the basis of the fed-in data and context information such as regional properties. The user can then still adjust this if necessary.During the simulation, the physical user interacts with these avatars acting within the simulation module according to their assigned roles, argument or respond to other participants. This allows a realistic replication of the negotiation without physical persons having to be present.These agent avatars may be presented in different ways, for example, depending on the requirements of the training scenario. One possibility is to display them visually on a screen or digital surface. This could be done in the form of a human-like representation, for example as an animated figure or as a photorealistic image of a person. As a result, the training user can interact with these visual representations in a similar manner to that in a real negotiation situation. For example, an avatar in a virtual training room could serve as a conversation partner, respond to questions, show mimic and gesture, or simulate various emotional responses.Alternatively, the interaction can also take place exclusively in the form of a text-based chat without a graphical or animated face being required. In such a case, the agent could only communicate with the user via text messages, similar to a classic chatbot. This type of text-based interaction may be particularly useful when focus is on voice communications, such as in practicing negotiation policies or simulating written communications in business or diplomatic contexts.In addition, hybrid approaches could be used in which both visual presentation and chat interaction are combined. For example, an avatar could appear on a screen while communication is via both voice and text messages. This could enhance the learning effect by using different communication channels in parallel, similar to a real conversation involving body language, tone case and word selection in common.In addition, the analysis module creates a negotiation simulation model on the basis of the retrieved data, which is supplied to the simulation module. Thus, for example, the specific sequence of a negotiation can be determined from the case data and integrated into the simulation model.A negotiation simulation model in this sense denotes a structured digital or software mapping of the sequence of a negotiation, which is created on the basis of relevant case data. It serves as a basis for the simulation of the method and determines how the various agent avatars act, argument and react to each other within the virtual negotiation. The model is created by the analysis module taking into account the retrieved data and then fed into the simulation module. It determines the specific structure and sequence of negotiation, including the order of talk contributions, interactions between participants, and possible decision processes. As a result, the negotiation simulation model enables realistic replication of negotiations that allows the user to train his negotiation capabilities effectively without having to have physical participants present.In the simulation module, the course of negotiation is thus simulated in a realistic manner, wherein the actions of the agent avatars are also taken into account. The physical user interacts directly with the system via the user interface, whereby the method can be simulated realistically without the need for further physical persons.The Kl-Bot module is, for example, a central component of the system for negotiation simulation and works closely together with the data module, the analysis module, the simulation module and the user interface, but it is not essential to the invention. Its primary object is to process the information stored and analyzed in the system, generate realistic interactions, and train the user's negotiation capabilities through dynamic, customizable simulations. In this case, the Kl-Bot module accesses the data module in which relevant case data are stored, fonts and, as described later, also argumentation patterns and method rules. This data uses it to develop realistic reasoning strategies and to react to negotiation dynamics in the simulation. In addition, the AI-Bot module interacts with the analysis module, which evaluates the case data retrieved and assigns specific properties and roles to the individual agent avatars. It supports the creation of the negotiation simulation model by recognizing patterns, calculating probabilities for specific negotiation behaviors, and simulating possible scenarios. Within the simulation module, the Kl-Bot module controls the behavior of the agent avatars and adjusts their responses to the user's interactions in real time. It simulates arguments, counter arguments, emotional responses, and tactics adjustments to create a highly realistic negotiation situation. In addition, it processes the findings obtained by the simulation and can further refine the negotiation scenarios based on the user behavior. Via the user interface, the AI-Bot module allows the user to interact directly with the system by analyzing his inputs, generating dynamic responses, and simulating the responses of other agents within the negotiation. In this case, it can evaluate the manner of argumentation of the user and suggest feedback or alternative strategies in a targeted manner to him. The Kl-Bot module thus acts as an intelligent control element within the system and orchestrations the various modules to enable a realistic negotiation simulation. In the broadest sense, the system thus has, for example, a AI-Bot module which processes data from the data module, cooperates with the analysis module for case analysis, influences the simulation module by the control of the agent avatars and interacts with the user via the user interface. Due to this tight networking, the system learns from each simulation and can continuously adapt to new negotiation scenarios.According to a further development of the invention, the data module can comprise a memory module in which the relevant case data are stored. This memory module can be supplied with data in particular via a data input interface. Via this data input interface, the relevant case data can be input into the data module.Relevant case data comprise e.g. information, in particular all information which is important for the simulation and analysis of a negotiation. These include, in particular, the written traffic from the method, the case description, the generation statements and agent reports created by persons involved in the method. These agent reports can be, for example, police accident reports or other official documentation which is stored in the method record and serves as the basis for the negotiation. Overall, relevant case data captures all available and essential information about a case that enables realistic and informed replication of the negotiation. These relevant case data can thus include, among other things, the case traffic from the method, the case description, the generation states or the agent reports. Agent reports are to be understood in this context as meaning that persons involved in the method have created reports which are stored in the file and are taken into account in the method. These may also be, for example, accident reports from the police. Preferably, all available relevant case data are read into the data module.This input can be made either by the user himself or by an authorized person in the attorney's plant.In addition, the data module can additionally have an interface to external systems (or else only to an external system), with the result that, in addition to the relevant case data, it also has access to external data. The external system may comprise one or more from the following group: a database with jurisdiction in the respective area, data generally available in the Internet, data with relevant information about the person of an agent, or a database with jurisdictions and / or laws in the corresponding jurisdiction. It is particularly advantageous if all the aforementioned external systems are connected to the data module via a corresponding interface, since a larger amount of information can thereby be incorporated into the negotiation.If it is a database with relevant information about the person of an agent, this can mean in the simplest case that, by knowing the name of a person, their social media accounts, such as, for example, Linkedln, are taken into account in order to construct a behavior and thinking pattern of the person in question therefrom. Ideally, data on the decisions of individual sheet testers are also available in order to understand their previous judgments, or data on the decisions of individual patent testers. Such information may be incorporated into the simulation and affect the behavior of the corresponding agent avatar, making the negotiation more realistic.The jurisdiction databases can be both known commercial providers and publicly available jurisdictions on the home pages of jurisdictions or offices.However, it is advantageous if the training user is informed, for example via the user interface, that such external information has entered the simulation and which specific data have been taken into account. As a result, the user himself can decide whether or not this information should be used in the simulation. After the user has received an appropriate message via the user interface, he could, for example, intentionally exclude the use of this data. This could be done, for example, by a pop-up window in which it has the option of confirming "OK" or "No".The external data are used in the simulation in a targeted manner, for example, by being fed into the data module and subsequently being processed by the analysis module. The aim is to simulate the negotiation as realistic as possible. Various external sources can be used, including jurisdiction databases, publicly available information, social media data, or specialized legal texts and legal precedence cases databases.The analysis module evaluates these data and assigns them to the respective agent avatars. These can thereby adapt their behavior, argumentation and response patterns based on real decision processes and background information. This contributes to the interaction not only taking into account realistic legal principles, but also dynamically entering into account specific developments and individual case constellations.An additional aspect is the ability to store a personality profile for each avatar to make the simulation even more authentic. The behavior of the respective agent is thereby based not only on facts, but also on individual properties. For example, publicly available information about a creator or negotiation partner could indicate that this person is extroversated, prone to emotional responses, or prefers a particularly relevant and analytical approach. Such features can then be incorporated into the simulation in a targeted manner, so that the avatar responds accordingly, for example by a pulsey expression or a retaining strategic argumentation.In addition, further context information can be included, which makes the simulation even more precise. The state of knowledge of the respective agent can be taken into account in this case, so that it occurs either as an expert with informed knowledge or as a lay-out with limited understanding for the topic. Relevant cases of precession can also be integrated, so that the system simulates realistic argumentation lines which rely on existing jurisdictions. Regional details may also play a role, for example by taking into account which legal or cultural factors are important for a specific negotiation situation.The combination of case-related data, individual personality profiles and contextual information makes the behavior of the agent avatars even more multi-layered and more adaptive. This enables realistic and interactive training scenarios in which the user has to adapt to different characters, knowledge levels and legal framework conditions.For example, if there is data on previous decisions of a director or patent verifier, the system may use this knowledge to account for predictable responses or argumentation patterns within the simulated negotiation. Likewise, analysis of social media profiles or other publicly available information may serve to mimic personality features and deciduous patterns of particular process participants. This allows the behavior of an agent avatar to be adapted to specific reaction modes in order to enable realistic interactions within the simulation.In addition, it can be ensured via the user interface that the training user is informed about the use of such external data. It receives a feedback about which external information has flowed into the simulation and can decide whether or not these should be taken into account. If the user wishes to exclude certain external data, he can do this, for example, via a corresponding option in a pop-up window. This dynamic inclusion of external data not only makes the simulation more adaptive, but can also develop more precise and individualized negotiation strategies by relying on infused external sources.According to a further development of the invention, the agents are selected from the following group: Judicative agent or Executive agent, agent of a first party of the method, agent of a second party of the method, agent of a first agent of the first party and agent of a second agent of the second party. The term judicative agent or executive agent is used herein as a generic term for director, publisher and other decision agents. A judicative agent or execitive agent can also be composed of a group of individual persons, for example, of director or patent checkers, who in the narrower sense do not have a judicative function. Therefore, they could also be referred to as executive agents because they are associated with an executive decision instance.Not only can two parties participate in the method, but any number of parties or even only a single party, as is the case, for example, in one-way methods in front of the patent office. In such a case, it suffices if at least one agent of a first party or one agent of a representative of the first party as well as a judicative agent are provided, wherein the latter can also be only a single person. However, an embodiment of the system is advantageous which enables the participation of at least three, preferably four, five, six, seven or even eight agents. This increases the complexity of the situation and requires a greater data processing capacity, which can be coped with accordingly, however.According to a further embodiment according to claim 6, the judicative agent can be generated from a plurality of individual persons of the board. In this case, in particular the relevant information available for the individual persons can be taken into account in order to generate a combined, averaged opinion therefrom, which then serves as a decision of the subject.For example, a corresponding application simulating the negotiation is stored in the simulation module. This application is special simulation software for simulating the negotiation. It is fed with the agent avatars as well as the relevant case data, in particular taking into account the established negotiation simulation model.Although the application is not claimed here independently, it can represent an independent invention on its own, so that special software is provided for simulating a negotiation within the method.This application takes into account the negotiation simulation model created by the analysis module and the agent avatars and outputs the corresponding simulation results to the user interface.The described application is thus a simulation software which serves to simulate a negotiation. It processes the relevant case data, controls the agent avatars and simulates the negotiation flow based on the negotiation simulation model. The application interacts with various components of the system by using the data and models created by the analysis module, transmitting them into the simulation module and outputting the results via the user interface.The data module supplies the relevant case data which are included in the simulation. The analysis module prepares this data, creates the negotiation simulation model, and assigns specific features and roles to the agent avatars. The simulation module takes control of the simulation and allows the agents to interact dynamically within the negotiation. Finally, the user interface ensures that the user can interact with the application, whether by visual display of the simulation, voice inputs or menu options for adapting the training parameters. Overall, the application ensures a realistic negotiation simulation by connecting the various system components to one another and outputting the simulation results to the user.The application and the AI-Bot module are both components of the system, but differ in their functions and task areas. The application is the master simulation software that controls the entire negotiation simulation by processing the relevant case data, negotiation simulation model, and agent avatars and outputting their interactions visually or auditively to the user interface. It forms the structural framework for the simulation and ensures that the sequence of the negotiation is adjusted according to the predefined data and rules.The Kl-Bot module, on the other hand, is a specialized component within this simulation, which is responsible for the dynamic, intelligent control of the agent avatars. It uses artificial intelligence to generate the arguments, behaviors, and responses of the simulated negotiation-related individuals in a realistic manner and adapt them to the inputs of the user. While the application ensures the technical flow of the simulation, the AI-BOT module functions as an intelligent decision and interaction entity, which allows agents to react flexibly to new arguments or strategies of the user.Moreover, the AI can make a prediction about the possible outputs of the negotiation based on the available data and the interactions performed in the simulation. By analyzing the previous decision processes, including relevant precedence cases, and recognizing patterns in the argumentation structures, the system can calculate probabilities for different negotiation behaviors. For example, it could be predicted that a director with a first probability of e.g. 93.4%, which is higher than the second probability, makes a certain decision, while alternative results are evaluated with the second probability of e.g. 5.2% or 1.4%. These predictions can help the user to better estimate the prospects of success of different reasoning strategies and to adapt his negotiation procedure accordingly.In summary, the application provides the platform and structure of the simulation, while the AI-BOT module ensures the content and interactive depth through realistic, adaptive behaviors of the agents.According to an advantageous development, the user interface can comprise a visual interface with a display screen, via which the user can select which aspects are to be trained with the system. This can be effected in particular via a dropdown menu. The user can thereby select which agent he takes over and wants to train in the simulation. Alternatively or additionally, it may determine which part of the negotiation it wishes to train or whether it wishes to simulate the entire negotiation.According to a further advantageous development, the user interface can be configured such that the user interacts with the system by voice at least during the training of the negotiation. This means that, in addition or as an alternative to the display screen, different system settings can be made via a dropdown menu, a speech-based interaction is possible during the training. The user interface can not only capture voice inputs of the user, but also output the voice of the various agents or display them as written text on the display screen. In addition, the different agents that are simulated in the negotiation can be assigned their own voice, which is then reproduced via the user interface.This makes it possible to use the system in a room that is simulated to a dish room, for example with a corresponding chair and screens on the walls on which the virtual avatars are visualized and interact with the user.According to an advantageous development of the system, the user interface has an interface via which documents can be input. This can simulate a written document being presented in the negotiation, for example a new request or a similar document. These documents are read in via the user interface or the corresponding data are transmitted to the system, so that the information is taken into account in the further negotiation.According to a further advantageous embodiment, the system is configured such that it detects the interaction with the user and takes into account the information obtained by inputting it for the further course of negotiation. As a result, the output and the behavior of the agents participating in the negotiation adapt dynamically, so that the simulation is continuously adapted to the respective situation.A further feature of the invention is the integration of the previously described Kl-Bot module, which serves for processing the data from the data module and the user inputs via the user interface. The Kl-Bot module, as described above, is a specialized component within an artificial intelligence stick that is responsible for certain functions or tasks. These modules cooperate to enable the bot to process natural language, make decisions, and respond to user requests.The system is further configured to simulate the negotiations in real-time and length. In particular, information about the prospective end of the negotiation can be provided via the user interface, wherein the current information position and the presented arguments are also taken into account. Depending on the argumentation, the prospective end of the negotiation may shift accordingly. This prognosis may be displayed on the visual display screen, for example.A further advantageous feature of the system is the ability to evaluate the effects of different argumentation strings of the training user in real time (instantan) and to provide a recommendation for a suitable strategy via the user interface. The user can try different argument strings, and for each of these strings a scenario is developed that evaluates the prospects of success. These evaluations are visually presented so that the user can decide informed which reasoning strategy is most promising.In addition, the system provides the ability to override any agent interaction and simulate alternative scenarios. The user can, for example, specifically examine how the further course and the result of the negotiation would change if a vehicle had made a different statement. Thus, the system could calculate on demand what effects it would have if a tool did not say "X... " but "Y... ". This allows the user to flexibly handle various variations of negotiation and analyze the consequences of different statements or arguments. This allows not only a deeper understanding of the negotiation dynamics, but also a targeted preparation for various possible developments within a case.According to a further advantageous embodiment, after completion of the training, the system can calculate the probability of success of the corresponding party and in particular visualize it on the display screen. On this basis, it can be decided whether participation in the negotiation is reasonable or whether it would be more economical to have the party decided in the absence to save cost.In addition, the system may include a translation model that transfers foreign language documents to the method language before being considered in the simulation. Since not all documents in the method are present in the method language, for example if a tenant is foreign-language, the translation model enables automatic processing of such documents so that they can be integrated into the simulation.According to a further advantageous embodiment, the system can have a simulation result storage module. This stores the course of the simulation including the relevant data as a file, so that the user can interrupt the negotiation simulation at any time and continue at a later time.The simulation result storage module serves, for example, to store the course of the negotiation simulation together with all relevant data in a file which can be stored and opened again and read in at a later point in time. As a result, the user can interrupt the simulation at any time and subsequently either retrieve it again for analysis of the previous course or continue the simulation at a specific point with changed arguments. In addition, the module makes it possible to continue an uncomputed negotiation simulation, so that the user can experiment flexibly with different strategies or complete the simulation.In addition, according to one refinement, the system is configured, for example, in such a way that it carries out probability analysis in the case of missing information, for example incomplete case data. On this basis, the missing information is replaced by the most likely values derived from the analysis, so that the simulation remains operable even in the case of incomplete inputs.The system is thus configured, for example, to remain functional even if certain information is missing, in particular if the relevant case data is incomplete. To ensure this, the system performs this probability analysis, in which existing data is evaluated and statistical models are used to replace the missing information with the most likely values. For example, the system takes into account comparable case patterns, previous decisions, or commonly available data sources to make an informed assessment of which information would be plausible. On this basis, the determined values are integrated into the simulation, so that the negotiation can still be simulated realistically, even if not all relevant data are present. This ensures that the simulation is not interrupted or falsified by incomplete inputs, but rather can be continued as close to reality as possible.Although the system itself is claimed herein, the invention is not limited to the system only. Rather, a computer program product or storage medium containing executable instructions for simulating a negotiation may also constitute a distinct invention. Likewise, although not claimed in the present application, a method for simulating a negotiation may be subject matter of the invention. In this case, the device-related aspects would be described in procedural form. Relevant case data would be retrieved in this case, assigned to the agents of the negotiation and agent avatars generated on the basis thereof. Subsequently, or before or simultaneously, a negotiation simulation model would be created and, based thereon, the negotiation would be simulated, enabling interaction with a training user.Further advantageous embodiments of the invention are illustrated in the following drawings.In these show: FIG. 1 illustrates setup of an exemplary system for simulating a negotiation; and FIG. 2 is a flow chart of how the relevant case data is assigned to the corresponding agent.Fig. 1 shows, with reference numeral 1, an input layer corresponding to the data module according to the claims. The system also comprises a core system, which is identified by reference numeral 2. Within this core system there are an analysis module, designated by reference numeral 3 as "analysis and planning", and a simulation module 4.The data module 1 contains a storage module 6, which is referred to as a document database in FIG. 1. In this memory module, for example, the relevant case data are stored, including the case traffic from the method, the case description, the generation statements and agent reports. These relevant case data are input via an input interface 7 which transfers the data into the memory module 6. In addition, the data module 1 has an interface to external systems, which is designated by reference numeral 8. This interface makes it possible to include external data sources in the simulation. These include, for example, legal databases with information relating to jurisdiction, publicly available Internet databases, databases with specific information relating to persons involved, and databases with legal texts or legal regulations in the corresponding jurisdiction. The system can be flexibly extended with further external data sources by this interface. The external data as well as the relevant case data stored in the memory module 6 are processed and taken into account in the core system 2. Based on these data, the simulation takes place within the simulation module 4.The simulation module 4 contains a simulation application 9 (which is described as an "application" in the general description) which is responsible for carrying out the simulation. The analysis module 3 has two central tasks: on the one hand, it analyzes the data retrieved from the data module 1 and assigns these to the respective agent, so that agent avatars are generated for the simulation.These avatars represent, for example, the dish, the counterpage, the defense and the vehicles. In this case, external information, such as personal data about an agent, can also be included in the simulation.The analysis module also establishes a negotiation simulation model that is taken into account in the simulation planning and integrated into the simulation application 9. This model defines the sequence and structure of the simulated negotiation by defining the negotiation strategy, argumentation sequences, and possible scenarios.While four agents are shown in Figure 1, the system may operate with any number of agents, requiring at least two agents. The simulation is not limited to a dish scenario, but may also include expecitive or judicative agents, depending on the type of negotiation. The system allows not only two parties to be included, but also to simulate one-way methods or negotiations with multiple participants. The created agent avatars are actively integrated into the process simulation and dynamically respond to the user's interactions.A Kl-Bot module, not shown, allows realistic interaction between the user and the simulated agent avatars. The Kl-Bot module processes information from the data module 1 and the user input via the user interface 5 to generate smart responses, argumentation patterns, and dynamic adjustments within the simulation. It can also supplement missing information based on probability analyses and comparable cases and evaluate alternative negotiation scenarios, for example.The user interface 5 allows the user to interact directly with the system. It may be a PC with display screen which can also be operated remotely in a different room than the core system 2. The system is designed such that the modules 1, 2 and 5 can also be geographically far apart from one another, for example on different continents. The display screen may include a drop-down menu through which the user may select which aspects of the system are to be trained. In this case, he can decide which agent he wishes to take over, whether he wishes to simulate a specific section or the entire negotiation and which strategies he wishes to test.Additionally, the user interface 5 may include a document entry capability to simulate a written document being presented during negotiation. This may be, for example, a request or evidence. The input documents are processed in the system and can influence the course of the simulation.A further advantageous embodiment provides that the system enables a speech-based interaction. During training, the user can interact with the system by voice and the agent avatars can render their statements either via voice output or in written form on the screen. The different agents can also receive different voices in order to ensure a simulation that is as realistic as possible. This can be used in particular in a room which is simulated to a dish room, for example with corresponding seating and screens on the walls on which the virtual avatars are projected.FIG. 2 illustrates the flow of data of the input relevant case data.In step S 1, these are uploaded into the system. Then, in step S2, the simulation planning takes place in the analysis module 3, the relevant case data being assigned to the corresponding agent. In this case, the system can be configured such that it is specified which data are accessible to which agent, so that only these data are taken into account for the simulation.In addition, the AI model can automate this process by independently analyzing which information is relevant to and allocating it accordingly to which agents. By machine learning and pattern recognition, the AI can automatically decide which data should be accessible to certain negotiation roles and which should not. This ensures that each simulated party only has the information that it would know in a real negotiation situation as well. This allows for even more efficient preparation of the simulation and significantly reduces the manual effort for data entry and assignment.The system is also designed to allow real-time negotiation simulation. It can display information about the prospective end of the negotiation to the user, wherein this is adapted based on the current information position and the arguments presented, it can also display information about the prospective result of the negotiation, e.g. with corresponding probabilities for a result. The simulation can evaluate different argument strings and make suggestions for a promising strategy for the user via the user interface. In addition, after completion of the simulation, the system can calculate and visualize a probability assessment for the success of the respective party, so that the user can decide whether participation in the real negotiation is appropriate.Further, the system may include a translation model that transfers foreign language documents to the method language before being considered in the simulation. This is particularly relevant when tenants are foreign-speaking and their inputs must be processed for the simulation.Moreover, the system may include a simulation result storage module storing the course of the simulation in a file. This file can be stored at any time, later re-opened and read in to analyze the simulation, continue with new arguments at a particular location, or continue an interrupted negotiation.If certain information is missing, for example relevant case data is incomplete, the system may perform probability analysis. In this case, the missing information is replaced by the most likely values that were derived from the analysis, so that the simulation continues to remain realistic and functional.Although the present invention has been described with reference to the specific example, it is not limited to the situation in right-handed use, but can be used for any negotiation in one-way or two-way methods both before expecitive organs and before judicative organs.While the system is claimed herein, the invention is not limited thereto. A computer program product or a medium with executable instructions for simulating a negotiation can also be the subject matter of the invention. A method for simulating a negotiation could likewise be claimed, in which relevant case data are retrieved, agent avatars are generated and a negotiation simulation model is set up in order to carry out an interactive simulation with the user.List of reference characters1 Input layer (data module) 2 Core system 3 Analysis module (analysis and planning) 4 Simulation module 5 User interface 6 Storage module (document database) 7 Input interface 8 Interface to external systems 9 Simulation application

Claims

Computer-implemented system for simulating a negotiation in a method, in particular jurisdiction negotiation, comprising a data module (1) for storing relevant case data; an analysis module (3) configured to assign the data retrieved from the data module (1) to agents of the negotiation in order to generate a respective agent avatar therefrom and to build a negotiation simulation model based on the retrieved data; a simulation module (4) simulating the negotiation with the aid of the negotiation simulation model and the agent avatars, and a user interface (5) via which a user to be trained interacts with the system as an agent of the negotiation.System according to Claim 1, characterized in that the data module (1) comprises a memory module (6) in which the relevant case data are stored, and wherein in particular a data input interface (7) is provided via which the relevant case data are input into the data module (6).The system according to claim 2, characterized in that the relevant case data is one or more or all selected from the group consisting of: written traffic from the method, case description, generate statements, agent reports on the case.The system according to any one of claims 1 to 3, characterized in that the data module (1) comprises an interface (8) to an external system from which corresponding data are retrieved, wherein the external system is one or more selected from the group: database with jurisdiction in the corresponding area, data generally available in the Internet, database with relevant information about the person of an agent, database with jurisdictions and / or laws in the corresponding area.The system according to any one of claims 1 to 4, characterized in that the agents are selected from the group consisting of: Judicative agent, Executive agent, First party agent of the method, Second party agent, Generate agent, First party agent, Second party agent.System according to claim 5 as dependent on claim 4, characterised in that the judicative agent or the executive agent is generated from several individual persons of the board, and in particular takes into account the relevant information on the person available to the individual persons in order to form a single averaged opinion for the judicative agent or executive agent thereon.System according to one of Claims 1 to 6, characterized in that the simulation module (4) stores an application which simulates the negotiation.The system according to any one of claims 1 to 7, characterized in that the user interface (5) is a visual user interface with a display screen via which the user can select, in particular can select in a drop-down menu, which aspects are to be trained with the system, in particular which agent is the user and wishes to train and / or which part of the negotiation or the entire one of them wishes to train.The system according to any of claims 1 to 8, characterized in that the user interface (5) is configured such that the user interacts with the system at least during training of the negotiation by voice.A system according to any one of claims 1 to 9, characterized in that the user interface (5) comprises an interface by means of which documents can be input in order to simulate the written document being input in the negotiation.A system according to any one of claims 1 to 10, characterized in that the system is configured to take into account the interaction with the user and the information of the user's input for the further negotiation history and to adapt the output and or the behavior of the agents participating in the negotiation thereto.System according to one of Claims 1 to 11, characterized in that a AI-Bot module is provided for processing the information from the data of the data module (1) and that of the input of the user via the user interface (5).The system according to any one of claims 1 to 12, characterized in that the system is configured to simulate the negotiation faithfully and in real time, wherein in particular an anticipated end of the negotiation at the current information location is made available to the user by means of the user interface.The system according to any one of claims 1 to 13, characterized in that the system is configured to evaluate the effects of different argumentation strings of the training user instantaneously and to specify a recommendation of a strategy via the user interface.The system of any of claims 1 to 14, characterized in that the system is configured to, after performing the training, indicate a probability of success for one of the parties.The system according to any one of claims 1 to 15, characterized in that the system comprises a translator module that translates foreign language documents into the method language before being considered in the system.System according to one of Claims 1 to 16, characterized in that the system has a simulation result storage module in which the profile together with data of the simulation carried out is stored as a file, it being possible in particular for these files, if the simulation has not been carried out until the end, to be read back into the system in order to continue the simulation.The system according to any one of claims 1 to 17, characterized in that the system is configured to perform probability analysis in the absence of information and then to take into account the most likely information in the simulation for the missing information.

Images