Pernille Hemmer, Ph.D. - US grants

The objective of this research is to examine how an individual's experiences influence his or her beliefs and how these, in turn, impact intertemporal decisions. Understanding how experiences and resulting beliefs impact future choice contributes to the national health through application to questions of medication adherence and health outcomes (e.g., how patient' uncertainty about illnesses might affect adherence choices.) The project also contributes to national prosperity and welfare through the provision of integrated in-class and laboratory training of graduate students and undergraduates in advanced programming and Bayesian Cognitive Modeling. The research program also seeks to advance diversity in STEM fields through outreach focused on attracting women and underrepresented groups to the academic setting, and continued post-degree mentoring, e.g., outreach to pre-college organizations serving underrepresented minorities and organizations promoting retention of women in science.

Our beliefs about the regularities of our environment shape our understanding of the world, and in turn, influence our cognitive processes and behavior. Beliefs are important not only for recalling the past and making decisions in the now, but also for making predictions about the future. People develop well-calibrated beliefs based on their life experiences. However, subjective experiences leading to individual differences in beliefs have also been found to bias decision processes. Time preference (placing greater value on earlier outcomes by discounting the utility of later outcomes) is one example where human choice and behavior is influenced by strong individual differences, as well as changes within the individual. While Discounted Utility, the accepted model of normative intertemporal choice, assumes consistent assignment of discounting rates in time preference (i.e., stable individual difference), it has been suggested that people might exhibit a change in time preference due to beliefs about the uncertainty in the environment. An important question that remains unexplored is: what is the effect of changes in the environment on subjective beliefs and the influence of changing beliefs on memory and decision making under uncertainty (e.g., predictions for the future). This research provides a new framework for inferring individual differences via the integrated application of Bayesian analysis to Bayesian models of cognition. This integrative approach makes it possible to infer individual differences in the underlying parameters of the Bayesian cognitive model, and can be used to address how changing beliefs influence how and what we remember, and the value of future outcomes. The theory driven goal is to challenge the assumption of a constant discounting rate, and instead assume that the rate changes, but that the system for determining the rate remains consistent. The integrative approach makes it possible to implement and compare multiple competing cognitive models of the effect of individual discounting rates on time preference.

This multidisciplinary project undertakes a program of research in natural language generation (NLG), the subfield of artificial intelligence that aims to construct intuitive, accessible utterances to communicate the data, knowledge and reasoning of computational systems. NLG capabilities have an important role in facilitating new, more natural interaction with computers, both in current applications such as mobile information access and in emerging ones such as personal assistants and human-robot interaction. NLG systems remain inflexible and difficult to build, however. This research aims to addresses this problem by developing techniques to train NLG systems to match human language use. The project is a close collaboration that links psychological experiments, designed to uncover the strategies human speakers use, to computational experiments, which apply these strategies in NLG systems using machine learning.

The theoretical framework at the center of this project is Bayesian cognitive modeling, a probabilistic approach that explains human information processing in terms of decision making under uncertainty. Applied to language use, Bayesian cognitive modeling involves estimating the communicative goals speakers adopt, the knowledge and meanings available to speakers, and the choices speakers make to express needed information in suitable linguistic terms. Such knowledge and strategies can then be used to drive NLG systems. The specific research of the project investigates three key domains for applying NLG to construct messages to describe real-world situations: making lexical choices, constructing complex linguistic structures compositionally, and fulfilling multiple overlapping communicative goals. The project explores each domain through interrelated activities carried out by an interdisciplinary team of computer scientists and psychologists: to formalize speaker choices using a range of Bayesian cognitive models; to fit the models to visually-grounded language corpora using machine learning; to evaluate the empirical scope of goal-directed reasoning by comparing the learned models both to attested human choices and to baseline learned models; and to assess how well the models match human comprehension of linguistic meaning. The intellectual merits of the project lie in bridging the gap between traditional goal-directed rational models of human behavior and state-of-the-art computational methods that instantiate templates or reproduce likely patterns. In addition to the societal impacts of the technology, the broader impacts of the project include the construction of data resources, models and modeling tools that will be distributed to facilitate further research, and contributions to ongoing initiatives for education in cognitive science at Rutgers.

This project examines terrorist attacks by individuals who are outside of an organized terrorist group (lone actor attackers) that target public spaces like train stations (soft targets). Such attacks have increased 134 percent in the last 20 years, yet lone actor attack-defend models have not kept up with the trend. The project will develop new models based on game-theory to understand attack and defense strategies combined with immersive simulations that can validate the theoretical models. The project team has expertise in the fields of operations research, industrial and systems engineering, psychology, and electrical and computer engineering. Implementation of this work will contribute to the national priority to reduce risk to critical infrastructures and their users. Furthermore, it will provide advanced training in game-theoretic models for undergraduate and graduate students. This scientific research contribution thus supports NSF's mission to promote the progress of science and to advance our national welfare. In this case, the benefits will be insights to improve man-made emergency management, which can save lives in future events.

This project addresses the gaps in current understanding of lone actor attacks to guide the development of new innovative defense strategies. Specific research objectives are: 1) to develop and analyze game-theoretic models of attack and defense strategies, and protection algorithms, to be used by the defenders against lone actor attackers; 2) to design immersive simulations to provide descriptive agents' behavior and to validate the game-theoretic models using risk metrics such as expected damage, and the fraction of unsuccessful attacks. The intellectual merit of this research is the broadening of the knowledge base of game theory with incomplete information, multi-agent (attacker and defender) learning, and stochastic games of partially observable systems. This is transformational research since it brings a fresh vision into the risk management, immersive simulations, statistical learning and normative behavior studies for infrastructure security. The anticipated results of this research are both analytical and practical for emergency management agencies, transportation safety officers, and the police.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.