Avi Rosenfeld

A study of mechanisms for improving robotic group performance

Many collaborative multi-robot application domains have limited areas of operation that cause spatial conflicts between robotic teammates. These spatial conflicts can cause the teams productivity to drop with the addition of robots. This phenomenon is impacted by the coordination methods used by the team-members, as different coordination methods yield radically different productivity results. However, selecting the best coordination method to be used by teammates is a formidable task. This paper presents techniques for creating adaptive coordination methods to address this challenge. We first present a combined coordination cost measure, CCC, to quantify the cost of group interactions. Our measure is useful for facilitating comparison between coordination methods, even when multiple cost factors are considered. We consistently find that as CCC values grow, group productivity falls. Using the CCC, we create adaptive coordination techniques that are able to dynamically adjust the efforts spent on coordination to match the number of perceived coordination conflicts in a group. We present two adaptation heuristics that are completely distributed and require no communication between robots. Using these heuristics, robots independently estimate their combined coordination cost (CCC), adjust their coordination methods to minimize it, and increase group productivity. We use simulated robots to perform thousands of experiment trials to demonstrate the efficacy of our approach. We show that using adaptive coordination methods create a statistically significant improvement in productivity over static methods, regardless of the group size.

Combining psychological models with machine learning to better predict people's decisions

Creating agents that proficiently interact with people is critical for many applications. Towards creating these agents, models are needed that effectively predict people’s decisions in a variety of problems. To date, two approaches have been suggested to generally describe people’s decision behavior. One approach creates a-priori predictions about people’s behavior, either based on theoretical rational behavior or based on psychological models, including bounded rationality. A second type of approach focuses on creating models based exclusively on observations of people’s behavior. At the forefront of these types of methods are various machine learning algorithms.This paper explores how these two approaches can be compared and combined in different types of domains. In relatively simple domains, both psychological models and machine learning yield clear prediction models with nearly identical results. In more complex domains, the exact action predicted by psychological models is not even clear, and machine learning models are even less accurate. Nonetheless, we present a novel approach of creating hybrid methods that incorporate features from psychological models in conjunction with machine learning in order to create significantly improved models for predicting people’s decisions. To demonstrate these claims, we present an overview of previous and new results, taken from representative domains ranging from a relatively simple optimization problem and complex domains such as negotiation and coordination without communication.

Modeling agents based on aspiration adaptation theory

Creating agents that realistically simulate and interact with people is an important problem. In this paper we present strong empirical evidence that such agents should be based on bounded rationality, and specifically on key elements from Aspiration Adaptation Theory (AAT). First, we analyzed the strategies people described they would use to solve two relatively basic optimization problems involving one and two parameters. Second, we studied the agents a different group of people wrote to solve these same problems. We then studied two realistic negotiation problems involving five and six parameters. Again, first we studied the negotiation strategies people used when interacting with other people. Then we studied two state of the art automated negotiation agents and negotiation sessions between these agents and people. We found that in both the optimizing and negotiation problems the overwhelming majority of automated agents and people used key elements from AAT, even when optimal solutions, machine learning techniques for solving multiple parameters, or bounded techniques other than AAT could have been implemented. We discuss the implications of our findings including suggestions for designing more effective agents for game and simulation environments.

A Study of Scalability Properties in Robotic Teams

In this chapter we describe how the productivity of homogeneous robots scales with group size. Economists found that the addition of workers into a group results in their contributing progressively less productivity; a concept called the Law of Marginal Returns. We study groups that differ in their coordination algorithms, and note that they display increasing marginal returns only until a certain group size. After this point the groups’ productivity drops with the addition of robots. Interestingly, the group size where this phenomenon occurs varies between groups using differing coordination methods. We define a measure of interference that enables comparison, and find a high negative correlation between interference and productivity within these groups. Effective coordination algorithms maintain increasing productivity over larger groups by reducing the team’s interference levels. Using this result we are able to examine the productivity of robotic groups in several simulated domains in thousands of trials. We find that in theory groups should always add productivity during size scale-up, but spatial limitations within domains cause robots to fail to achieve this ideal. We believe that coordination methods can be developed that improve a group’s performance by minimizing interference. We present our findings of composite coordination methods that provide evidence of this claim.

Learning Drivers’ Behavior to Improve Adaptive Cruise Control

Traditionally, vehicles have been considered as machines that are controlled by humans for the purpose of transportation. A more modern view is to envision drivers and passengers as actively interacting with a complex automated system. Such interactive activity leads us to consider intelligent and advanced ways of interaction leading to cars that can adapt to their drivers. In this article, we focus on the adaptive cruise control (ACC) technology that allows a vehicle to automatically adjust its speed to maintain a preset distance from the vehicle in front of it based on the driver’s preferences. Although individual drivers have different driving styles and preferences, current systems do not distinguish among users. We introduce an approach to combine machine learning algorithms with demographic information and behavioral driver models into existing automated assistive systems. This approach can reduce the interactions between drivers and automated systems by adjusting parameters relevant to the operation of these systems based on their specific drivers and context of drive. We also learn when users tend to engage and disengage the automated system. This approach sheds light on the kinds of dynamics that users develop while interacting with automation and can teach us how to improve these systems for the benefit of their users. While generic packages such as Weka were successful in learning drivers’ behavior exclusively based on the ACC’s sensors, we found that improved learning models could be developed by adding information on drivers’ demographics and a previously developed model about different driver types. We present the general methodology of our learning procedure and suggest applications of our approach to other domains as well.

NegoChat-A: a chat-based negotiation agent with bounded rationality

To date, a variety of automated negotiation agents have been created. While each of these agents has been shown to be effective in negotiating with people in specific environments, they typically lack the natural language processing support required to enable real-world types of interactions. To address this limitation, we present NegoChat-A, an agent that incorporates several significant research contributions. First, we found that simply modifying existing agents to include an natural language processing module is insufficient to create these agents. Instead, agents that support natural language must have strategies that allow for partial agreements and issue-by-issue interactions. Second, we present NegoChat-A’s negotiation algorithm. This algorithm is based on bounded rationality, and specifically anchoring and aspiration adaptation theory. The agent begins each negotiation interaction by proposing a full offer, which serves as its anchor. Assuming this offer is not accepted, the agent then proceeds to negotiate via partial agreements, proposing the next issue for negotiation based on people’s typical urgency, or order of importance. We present a rigorous evaluation of NegoChat-A, showing its effectiveness in two different negotiation roles.

Predicting and Identifying Missing Node Information in Social Networks

In recent years, social networks have surged in popularity. One key aspect of social network research is identifying important missing information that is not explicitly represented in the network, or is not visible to all. To date, this line of research typically focused on finding the connections that are missing between nodes, a challenge typically termed as the link prediction problem. This article introduces the missing node identification problem, where missing members in the social network structure must be identified. In this problem, indications of missing nodes are assumed to exist. Given these indications and a partial network, we must assess which indications originate from the same missing node and determine the full network structure. Toward solving this problem, we present the missing node identification by spectral clustering algorithm (MISC), an approach based on a spectral clustering algorithm, combined with nodes’ pairwise affinity measures that were adopted from link prediction research. We evaluate the performance of our approach in different problem settings and scenarios, using real-life data from Facebook. The results show that our approach has beneficial results and can be effective in solving the missing node identification problem. In addition, this article also presents R-MISC, which uses a sparse matrix representation, efficient algorithms for calculating the nodes’ pairwise affinity, and a proprietary dimension reduction technique to enable scaling the MISC algorithm to large networks of more than 100,000 nodes. Last, we consider problem settings where some of the indications are unknown. Two algorithms are suggested for this problem: speculative MISC, based on MISC, and missing link completion, based on classical link prediction literature. We show that speculative MISC outperforms missing link completion.

PHIRST: A distributed architecture for P2P information retrieval

Recent progress in peer to peer (P2P) search algorithms has presented viable structured and unstructured approaches for full-text search. We posit that these existing approaches are each best suited for different types of queries. We present PHIRST, the first system to facilitate effective full-text search within P2P databases. PHIRST works by effectively leveraging between the relative strengths of these approaches. Similar to structured approaches, agents first publish terms within their stored documents. However, frequent terms are quickly identified and not exhaustively stored, resulting in a significant reduction in the systems storage requirements. During query lookup, agents use unstructured search to compensate for the lack of fully published terms. Additionally, they explicitly weigh between the costs involved in structured and unstructured approaches, allowing for a significant reduction in query costs. Finally, we address how node failures can be effectively addressed through storing multiple copies of selected data. We evaluated the effectiveness of our approach using both real-world and artificial queries. We found that in most situations our approach yields near perfect recall. We discuss the limitations of our system, as well as possible compensatory strategies.

Classifying Papers from Different Computer Science Conferences

This paper analyzes what stylistic characteristics differentiate different styles of writing, and specifically types of different A-level computer science articles. To do so, we compared various full papers using stylistic feature sets and a supervised machine learning method. We report on the success of this approach in identifying papers from the last 6 years of the following three conferences: SIGIR, ACL, and AAMAS. This approach achieves high accuracy results of 95.86%, 97.04%, 93.22%, and 92.14% for the following four classification experiments: (1) SIGIR / ACL, (2) SIGIR / AAMAS, (3) ACL / AAMAS, and (4) SIGIR / ACL / AAMAS, respectively. The Part of Speech (PoS) and the Orthographic sets were superior to all others and have been found as key components in different types of writing.

Group planning with time constraints

Embedding planning systems in real-world domains has led to the necessity of Distributed Continual Planning (DCP) systems where planning activities are distributed across multiple agents and plan generation may occur concurrently with plan execution. A key challenge in DCP systems is how to coordinate activities for a group of planning agents. This problem is compounded when these agents are situated in a real-world dynamic domain where the agents often encounter differing, incomplete, and possibly inconsistent views of their environment. To date, DCP systems have only focused on cases where agents’ behavior is designed to optimize a global plan. In contrast, this paper presents a temporal reasoning mechanism for self-interested planning agents. To do so, we model agents’ behavior based on the Belief-Desire-Intention (BDI) theoretical model of cooperation, while modeling dynamic joint plans with group time constraints through creating hierarchical abstraction plans integrated with temporal constraints network. The contribution of this paper is threefold: (i) the BDI model specifies a behavior for self interested agents working in a group, permitting an individual agent to schedule its activities in an autonomous fashion, while taking into consideration temporal constraints of its group members; (ii) abstract plans allow the group to plan a joint action without explicitly describing all possible states in advance, making it possible to reduce the number of states which need to be considered in a BDI-based approach; and (iii) a temporal constraints network enables each agent to reason by itself about the best time for scheduling activities, making it possible to reduce coordination messages among a group. The mechanism ensures temporal consistency of a cooperative plan, enables the interleaving of planning and execution at both individual and group levels. We report on how the mechanism was implemented within a commercial training and simulation application, and present empirical evidence of its effectiveness in real-life scenarios and in reducing communication to coordinate group members’ activities.