Automated negotiation, a form of interaction among autonomous agents, plays a central role in multi-agent systems, yet the application of large language model (LLM) in this domain remains underexplored. An LLM can serve as a meta-strategist, adaptively selecting explicit strategies for execution by external strategic tools based on its capabilities. We propose a negotiation AI agent equipped with explicit strategic tools, including time-dependent and tit-for-tat negotiation strategies. Our results show that strategic tool enhanced negotiators achieve approximately 16% higher average utility compared with baseline, latest LLM negotiators.

Although centralized training with centralized execution (CTCE) excels at multi-agent coordination, its reliance on global information limits its use in the real world. Conversely, the practical decentralized execution (CTDE) paradigm often struggles with complex coordination. This paper bridges this critical gap by introducing the Centralized-to-Decentralized (CtoD) learning concept: a novel framework for transferring the knowledge of a powerful centralized policy into a robust, practical decentralized policy. Our method, CtoD-MAT, realizes this transition through a curriculum that gradually shifts agents from centralized to decentralized control. A key innovation is our dynamic scheduling mechanism, featuring a mediator module, which ensures a robust and effective knowledge transfer. Using challenging SMAC benchmarks, we demonstrate that CtoD-MAT successfully produces competitive decentralized policies, notably solving complex coordination tasks that are difficult for standard CTDE methods.

Multi-agent reinforcement learning (MARL) trains multiple agents in shared environments. Recently, MARL models have significantly improved performance by leveraging sequential decision-making processes. Although these models can enhance performance, they do not explicitly con-sider the importance of the order in which agents make decisions. We propose AOAD-MAT, a novel model incorporating action decision sequence into learning. AOAD-MAT uses a Transformer-based actor-critic architecture to dynamically adjust agent action order. It introduces a subtask predicting the next agent to act, integrated into a PPO-based loss function. Experiments on StarCraft Multi-Agent Challenge and Multi-Agent MuJoCo benchmarks show AOAD-MAT out-performs existing models, demonstrating the effectiveness of adjusting agent order in MARL.

We propose an automated negotiation for a reinforcement learning agent to adapt the agent to unexpected situations such as demand changes in supply chain management (SCM). Existing studies that consider reinforcement learning and SCM assume a centralized environment where the coordination of chain components is hierarchical rather than through negotiations between agents. This study focused on a negotiation agent that considered the value function of reinforcement learning for SCM as its utility function in automated negotiation. We demonstrated that the proposed approach could avoid inventory shortages under increased demand requests from the terminal customer.

Recently, the damage to agriculture, forestry, and fishery by wildlife has become a serious social problem. The damage caused by a conflict between wildlife and human society often causes less motivation for farming and is expected to significantly inhibit the development of local agriculture and forestry. However, the current management of wildlife is limited to primitive methods such as stabilizing the population by culling reliant on experience. To prevent this problem, the trajectory prediction of wildlife is one of the solutions for efficient wildlife management. In this study, we propose a machine learning architecture for predicting wildlife trajectories, considering surrounding environmental factors. Particularly, we propose a machine learning architecture that more accurately predicts trajectories using a timeseries transformer model that can learn long-term dependencies and add a submodel that can consider surrounding environmental factors. The proposed architecture allows interpreting which past trajectory areas should be focused on when predicting a trajectory by visualizing the weight of the focus mechanism, one of the elements comprising the transformer model. Further, the proposed architecture can decide the priority of elements of a multivariate timeseries input by considering a variable selection network (VSN), which considers the relationship of timeseries between a given time point and those before and after using casual convolutions. The main advantage of our methods is to understand the potential environmental factors to predict animal trajectories by analyzing the importance of each input feature representation in VSN. We experimented to evaluate the proposed method using one of the largest biologging datasets of sika deer (Cervus nippon) in Kanagawa Prefecture, Japan. The proposed method outperforms existing methods on various evaluation metrics and can effectively consider environmental factors in predicting trajectories. In addition, we analyze the results and show that the proposed architecture can pay attention to the times that individuals moved significantly within the observation period by visualizing the trajectories for the best and worst cases and the weight of each environmental factor.

This study proposed a novel reward-based negotiating agent strategy using an issue-based represented deep policy network. We compared the negotiation strategies with reinforcement learning (RL) by the tournaments toward heuristics-based champion agents in multi-issue negotiation. A bilateral multi-issue negotiation in which the two agents exchange offers in turn was considered. Existing RL architectures for a negotiation strategy incorporate rich utility function that provides concrete information even though the rewards of RL are considered as generalized signals in practice. Additionally, in existing reinforcement learning architectures for negotiation strategies, both the issue-based representations of the negotiation problems and the policy network to improve the scalability of negotiation domains are yet to be considered. This study proposed a novel reward-based negotiation strategy through deep RL by considering an issue-based represented deep policy network for multi-issue negotiation. Comparative studies analyzed the significant properties of negotiation strategies with RL. The results revealed that the policy-based learning agents with issue-based representations achieved comparable or higher utility than the state-of-the-art baselines with RL and heuristics, especially in the large-sized domains. Additionally, negotiation strategies with RL based on the policy network can achieve agreements by effectively using each step.

Previous research on the comprehensive negotiation strategy using deep reinforcement learning (RL) has scalability issues of not performing effectively in the large-sized domains. We improve negotiation strategy via deep RL by considering an issue-based represented deep policy network to deal with multi-issue negotiation. The architecture of the proposed learning agent considers the characteristics of multi-issue negotiation domains and policy-based learning. We demonstrate that proposed method achieve equivalent or higher utility than existing negotiation agents in the large-sized domains.

Existing research in the field of automated negotiation considers a negotiation architecture in which some of the negotiation components are designed separately by reinforcement learning (RL), but comprehensive negotiation strategy design has not been achieved. In this study, we formulated an RL model based on a Markov decision process (MDP) for bilateral multi-issue negotiations. We propose a versatile negotiating agent that can effectively learn various negotiation strategies and domains through comprehensive strategies using deep RL. We show that the proposed method can achieve the same or better utility than existing negotiation agents.

Thanks to the success of goal-oriented negotiation dialogue systems, studies of negotiation dialogue have gained momentum in terms of both human-human negotiation support and dialogue systems.However, the field suffers from a paucity of available negotiation corpora, which hinders further development and makes it difficult to test new methodologies in novel negotiation settings.Here, we share a human-human negotiation dialogue dataset in a job interview scenario that features increased complexities in terms of the number of possible solutions and a utility function.We test the proposed corpus using a breakdown detection task for human-human negotiation support.We also introduce a dialogue act-based breakdown detection method, focusing on dialogue flow that is applicable to various corpora.Our results show that our proposed method features comparable detection performance to text-based approaches in existing corpora and better results in the proposed dataset.

In human-human negotiation, reaching a rational agreement can be difficult, and unfortunately, the negotiations sometimes break down because of conflicts of interests. If artificial intelligence can play a role in assisting with human-human negotiation, it can assist in avoiding negotiation breakdown, leading to a rational agreement. Therefore, this study focuses on end-to-end tasks for predicting the outcome of a negotiation dialogue in natural language. Our task is modeled using a gated recurrent unit and a pre-trained language model: BERT as the baseline. Experimental results demonstrate that the proposed tasks are feasible on two negotiation dialogue datasets, and that signs of a breakdown can be detected in the early stages using the baselines even if the models are used in a partial dialogue history.

Persuasions are common in online arguments such as discussion forums. To analyze persuasive strategies, it is important to understand how individuals construct posts and comments based on the semantics of the argumentative components. In addition to understanding how we construct arguments, understanding how a user post interacts with other posts (i.e., argumentative inter-post relation) still remains a challenge. Therefore, in this study, we developed a novel annotation scheme and corpus that capture both user-generated inner-post arguments and inter-post relations between users in ChangeMyView, a persuasive forum. Our corpus consists of arguments with 4612 elementary units (EUs) (i.e., propositions), 2713 EU-to-EU argumentative relations, and 605 inter-post argumentative relations in 115 threads. We analyzed the annotated corpus to identify the characteristics of online persuasive arguments, and the results revealed persuasive documents have more claims than non-persuasive ones and different interaction patterns among persuasive and non-persuasive documents. Our corpus can be used as a resource for analyzing persuasiveness and training an argument mining system to identify and extract argument structures. The annotated corpus and annotation guidelines have been made publicly available.

This paper focuses on fundamental research that combines syntactic knowledge with neural studies, which utilize syntactic information in argument component identification and classification (AC-I/C) tasks in argument mining (AM). The following are our paper’s contributions: 1) We propose a way of incorporating a syntactic GCN into multi-task learning models for AC-I/C tasks. 2) We demonstrate the valid effectiveness of our proposed syntactic GCN in fair experiments in some datasets. We also found that syntactic GCNs are promising for lexically independent scenarios. Our code in the experiments is available for reproducibility.1

Gaku Morio, Ryo Egawa, Katsuhide Fujita. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

This demonstration paper presents an argument-inducing online forum that stimulates participants with lack of premises for their claim in online discussions. The proposed forum provides its participants the following two subsystems: (1) Argument estimator for online discussions automatically generates a visualization of the argument structures in posts based on argument mining. The forum indicates structures such as claim-premise relations in real time by exploiting a state-of-the-art deep learning model. (2) Argument-inducing agent for online discussion (AIAD) automatically generates a reply post based on the argument estimator requesting further reasons to improve the argumentation of participants.

Bilateral multi‐issue closed negotiation is an important class for real‐life negotiations. Usually, negotiation problems have constraints such as a complex and unknown opponent's utility in real time, or time discounting. In the class of negotiation with some constraints, the effective automated negotiation agents can adjust their behavior depending on the characteristics of their opponents and negotiation scenarios. Recently, the attention of this study has focused on the interleaving learning with negotiation strategies from the past negotiation sessions. By analyzing the past negotiation sessions, agents can estimate the opponent's utility function based on exchanging bids. In this article, we propose a negotiation strategy that estimates the opponent's strategies based on the past negotiation sessions. Our agent tries to compromise to the estimated maximum utility of the opponent by the end of the negotiation. In addition, our agent can adjust the speed of compromise by judging the opponent's Thomas–Kilmann conflict mode and search for the Pareto frontier using past negotiation sessions. In the experiments, we demonstrate that the proposed agent has better outcomes and greater search technique for the Pareto frontier than existing agents in the linear and nonlinear utility functions.

The annual International Automated Negotiating Agents Competition (ANAC) is used by the automated negotiation research community to benchmark and evaluate its work andto challenge itself. The benchmark problems and evaluation results and the protocols and strategies developed are available to the wider research community.

Most real‐world negotiation involves multiple interdependent issues, which makes an agent’s utility functions nonlinear. Traditional negotiation mechanisms, which were designed for linear utilities, do not fare well in nonlinear contexts. One of the main challenges in developing effective nonlinear negotiation protocols is scalability; they cannot find a high‐quality solution when there are many issues, due to computational intractability. One reasonable approach to reducing computational cost, while maintaining good quality outcomes, is to decompose the utility space into several largely independent subspaces. In this paper, we propose a method for decomposing a utility space based on every agent’s utility space. In addition, the mediator finds the contracts in each group based on the votes from all agents, and combines the contract in each issue‐group. This method allows good outcomes with greater scalability than the method without issue‐grouping. We demonstrate that our protocol, based on issue‐groups, has a higher optimality rate than previous efforts, and discuss the impact on the optimality of the negotiation outcomes.

There is an increasing interest in complex automated negotiations, where agents negotiate about multiple, interdependent issues and agent utility functions exhibit low autocorrelation. In these scenarios, the negotiation mechanisms used to find agreement solutions among agents tend to fail due to the complexity of agents’ preference spaces, and this tendency increases as the degree of autocorrelation decreases. In this paper, we propose an automated negotiation model specially tailored for highly uncorrelated utility spaces based on weighted constraints. The model relies on a mediated, auction‐based interaction protocol and a set of heuristic mechanisms for bidding and deal identification. To address the challenges raised by highly uncorrelated utility spaces, we propose to use a quality factor, which allows agents to balance utility and deal probability when placing their bids or when searching for agreement regions among these bids. Experiments show that the proposed negotiation model achieves high optimality results and low failure rates even in negotiation scenarios involving highly uncorrelated utility spaces, thus outperforming previous approaches.

Most real-world negotiation involves multiple interdependent issues, which makes an agent's utility functions complex. Traditional negotiation mechanisms, which were designed for linear utilities, do not fare well in nonlinear contexts. One of the main challenges in developing effective nonlinear negotiation protocols is scalability; it can be extremely difficult to find high-quality solutions when there are many issues, due to computational intractability. One reasonable approach to reducing computational cost, while maintaining good quality outcomes, is to decompose the contract space into several largely independent sub-spaces. In this paper, we propose a method for decomposing a contract space into sub-spaces based on the agent's utility functions. A mediator finds sub-contracts in each sub-space based on votes from the agents, and combines the sub-contracts to produce the final agreement. We demonstrate, experimentally, that our protocol allows high-optimality outcomes with greater scalability than previous efforts. We also address incentive compatibility issues. Any voting scheme introduces the potential for strategic non-truthful voting by the agents, and our method is no exception. For example, one of the agents may always vote truthfully, while the other exaggerates so that its votes are always "strong." It has been shown that this biases the negotiation outcomes to favor the exaggerator, at the cost of reduced social welfare. We employ the limitation of strong votes to the method of decomposing the contract space into several largely independent sub-spaces. We investigate whether and how this approach can be applied to the method of decomposing a contract space.