Measuring causality between collaborative and individual gaze metrics for collaborative problem-solving with intelligent tutoring systems

When students are working collaboratively and communicating verbally in a technology-enhanced environment, the system cannot track what collaboration is happening outside of the technology, making it difficult to fully assess the collaboration of the students and adapt accordingly. In this article, we propose using gaze measures as a proxy for cognitive processes to achieve collaboration awareness. Specifically, we use Granger causality to analyse the causal relationships between collaborative and individual gaze measures from students working on a fractions intelligent tutoring system and the influence that the students' dialogue, prior knowledge, or success has on these relationships. We found that collaborative gaze patterns drive the individual focus in the pairs with high posttest scores and when they are engaged in problem-solving dialogues but the opposite with low performing students. Our work adds to the literature by extending the correlational relationships between individual and collaborative gaze measures to causal relationships and suggests indicators that can be used within an adaptive system.

Psychobiological Vulnerability to Stress

Jorge Eduardo Castro Cifuentes, Maria del Carmen Sandi Perez

Several psychobiological factors are associated with vulnerability/resilience to stress-induced depression. Emerging evidence indicates behavioral traits, such as anxiety, can be related to depression-like symptoms after exposure to chronic stress. However, single traits cannot explain the variability observed in individual's vulnerability to stress, highlighting that a combination of behavioral traits might provide a better characterization of the individual's vulnerability to development of depression following prolonged stress. Hippocampal neurogenesis has been highlighted as a cellular mechanism involved in the reduction on hippocampal volume under conditions of stress and depression, suggesting that a lack of new cells and connections may underlie the cognitive and emotional symptoms of mood disorders, and modulating neurogenesis is regarded as a target/mechanism for antidepressants actions. Macrophage migration inhibitory factor (MIF) is a pro-inflammatory cytokine that is expressed and potentially involved in cellular proliferation in the central nervous system. MIF was shown to induce cell cycle progression via activation of the extracellular signal-regulated kinase (ERK) cascade. Stress has been shown to alter the degree of activation of the multifunctional ERK 1/2 -as indicated by its phosphorylation (p) pERK1/2- and pharmacological manipulations of the ERK pathway affected anxiety and coping behavior. Nonetheless, other mediating mechanisms are involved in the course of both depression and antidepressant actions. Even though the amygdala has been related to the etiology and recovery of depression through imaging studies showing correlations between these mood conditions and amygdala's volume and functional activation, the number of studies devoted to this brain area is still scarce. We hypothesize here an essential role of the amygdala on depression and antidepressant actions based on the implication of the basolateral complex (BLA) in emotional arousal and stress-induced modulation of cognitive processes, and on recent data showing activation of corticotropin releasing factor receptors (CRFR1) as a substrate for stress-induced alterations. Interestingly, alterations in the functional connectivity between the amygdala, the hippocampus and frontal areas have been identified in individuals with genetic vulnerability to depression. Looking for such psychobiological interactions is particularly important today because of increasingly incidence of depression and stress related disorders'. Identifying and understanding psychobiological vulnerability to stress is crucial for best oriented and more effective prevention and intervention strategies. Therefore, the aim of this thesis was: i) to identify in rats behavioral traits capable of predicting differences in vulnerability/resilience to the behavioral, endocrine, neurogenic and brain activity effects of stress, ii) to explore the role of the amygdala and personality traits in the antidepressants' effects on depression-like behavior and hippocampal neurogenesis, and iii) to study the role of MIF expression in hippocampal neurogenesis and anxiety- and depression- like behaviors. Our first study indicates that both the direction and intensity of the impact of stress on behavior corresponds to different patterns of neurogenesis impairment and brain activity; moreover, these effects are dependent on the interaction between personality traits and the length of stress exposure. Chronic stress induced progressively depressive-like symptoms associated with an increase in cell proliferation and a decrease of cell survival in the hippocampus and with a decrease in ERK1/2 activity in the prefrontal cortex, the hippocampus and a U-shape effect in the amygdala. Personality-like profiles of high exploration seem to play a resilient role for the development of stress-induced depressive-like symptoms and the combination of high anxiety and low exploration seems to lead to higher vulnerability to develop depression-like behaviors. Whereas sub chronic stress increased cell proliferation in the combined profile of high anxiety and low exploration, profiles of low exploration showed a decrease in cell survival after exposure to chronic stress. Interestingly, ERK1/2 activity was decreased in control animals of the personality profile high anxiety and low exploration and subsequently increased after sub-chronic and chronic stress in the paraventricular nucleus, the amygdala and the hippocampus but not in the orbitofrontal cortex. The second study provides evidence for an important role for the amygdala on fluoxetine-stimulated cell proliferation and survival, as well as on the establishment of a link between cell proliferation and depression-like behavior. We also show an important role for anxiety in mediating the effects of basolateral amygdala and antidepressant activity on cell proliferation and survival. Chronic fluoxetine treatment had a positive effect on hippocampal cell survival only when the BLA was lesioned. Anxiety was related to hippocampal cell survival in opposite ways in sham- and BLA-lesioned animals (i.e., negatively in sham- and positively in BLA-lesioned animals). Both BLA lesions and low anxiety were critical factors to enable a negative relationship between cell proliferation and depression-like behavior. Therefore, our study highlights a role for the amygdala on fluoxetine-stimulated cell survival and on the establishment of a link between cell proliferation and depression-like behavior. It also reveals an important modulatory role for anxiety on cell proliferation involving both BLA-dependent and -independent mechanisms. The third study demonstrates the relevance of MIF expression for adult hippocampal neurogenesis. We identify MIF expression in neurogenic cells (in stem cells, cells undergoing proliferation and in newly proliferated cells undergoing maturation) in the subgranular zone of the rodent dentate gyrus. A causal role for MIF in cell proliferation was demonstrated using genetic and pharmacological approaches. Behaviorally, genetic deletion of MIF resulted in increased anxiety- and depression-like behaviors, as well as of impaired hippocampus-dependent memory. Together, our studies provide evidence supporting a pivotal role for MIF in both basal and antidepressant-stimulated adult hippocampal cell proliferation. Moreover, loss of MIF results in a behavioral phenotype that to a large extent corresponds with alterations predicted to arise from reduced hippocampal neurogenesis. These findings underscore MIF as a potentially relevant molecular target for the development of treatments linked to deficits in neurogenesis, as well as to problems related to anxiety, depression and cognition. Altogether the results presented support the idea of high anxiety trait individuals being a particularly prone phenotype and exposure to stressful life events as a requirement for individuals with this personality trait to develop depression.

EPFL2012

It is not either or: An initial investigation into combining collaborative and individual learning using an ITS

Jennifer Kaitlyn Olsen

Research on Computer-Supported Collaborative Learning (CSCL) has provided significant insights into why collaborative learning is effective and how we can effectively provide support for it. Building on this knowledge, we can investigate when collaboration is beneficial to support learning. Specifically, collaborative and individual learning are often combined in the classroom, and it is important for the CSCL community to understand when a combination is beneficial compared to individual or collaborative learning alone. Before investing significant work into discovering these details, an initial investigation is needed to determine if there may be any value in a combination. In this study, we compared a combined condition to individual or collaborative-only learning conditions using an intelligent tutoring system for fractions. The study was conducted with 382 4th and 5th grade students. Students across all three conditions had significant learning gains, but the combined condition had higher learning gains than the other conditions. However, this difference was restricted to the 4th grade students. By analyzing the hints and errors of students over time from process data, we found that students in the combined condition tended to make fewer errors both when working collaboratively and individually, and asked for fewer hints than the students in the other conditions. Students who collaborated (collaborative and combined conditions) also reported having higher situational interest in the activity. By finding support for the effectiveness of combining collaborative and individual learning, this paper opens a broader line of inquiry into how they can effectively be combined to support learning.

2019

Scalable Multi-agent Coordination and Resource Sharing

Panayiotis Danassis

A plethora of real world problems consist of a number of agents that interact, learn, cooperate, coordinate, and compete with others in ever more complex environments. Examples include autonomous vehicles, robotic agents, intelligent infrastructure, IoT devices, and so on. As more and more autonomous agents are deployed in the real-world, it will bring forth the need for novel algorithms, theory, and tools to enable coordination on a massive scale. In this thesis, we develop such tools to tackle two central challenges in multi-agent coordination research: solving allocation problems, and resource sharing, focusing on solutions that are scalable, practical, and applicable to real-world problems.In the first part of the thesis we tackle the problem of allocating resources to agents, i.e., solving a weighted matching problem. Real-world matching problems may occur in massively large systems, they are distributed and information-restrictive, and individuals have to reveal their preferences over the possible matches in order to get a high quality match, which brings forth significant privacy risks. As such, there are three main challenges: complexity, communication, and privacy.Our proposed approach, ALMA, is a practical heuristic designed for real-world, large-scale (

10^6

agents) applications. It is based on a simple altruistic behavioral convention: agents have a higher probability to back-off from contesting a resource if they have good alternatives, potentially freeing the resource for some agent that does not. ALMA tackles all of the aforementioned challenges: it is decentralized, runs on-device, requires no inter-agent communication, converges in constant time -- under reasonable assumptions --, and provides strong, worst-case, privacy guarantees. Moreover, by incorporating learning we can mitigate the loss in social welfare and increase fairness. Finally, rational agents can use such simple conventions, along with an arbitrary signal from the environment, to learn a correlated equilibrium for accessing a set resources, under high congestion.In the second part of the thesis we focus on a critical open problem: the question of cooperation in socio-ecological and socio-economical systems, and sustainability in the use of common-pool resources. In recent years, learning agents, especially deep reinforcement learning agents, have become ubiquitous in such systems. Yet, scaling to environments with a large number of agents and low observability continues to be a challenge. In our work, we focus on common-pool resources. Individuals face strong incentives to appropriate, which results in overuse and even the depletion of the resources. Our goal is to apply simple interventions to steer the population to desirable states.We propose a simple, yet powerful, and robust technique: allow agents to observe an arbitrary common signal from the environment. The agents learn to couple their policies, and avoid depletion in a wider range of settings, while achieving higher social welfare and convergence speed.Finally, we propose a practical approach to computing market prices and allocations via a deep reinforcement learning policymaker agent. Compared to the idealized market equilibrium outcome -- which can not always be efficiently computed -- our policymaker is much more flexible, allowing us to tune the prices with regard to diverse objectives such as sustainability and resource wastefulness, fairness, buyers' and sellers' welfare, etc.

EPFL2022