Past Interactions Research Articles

Learning from past society-environment interactions is discursive rather than substantive

Abstract Environmental changes pose unprecedented risks to human societies worldwide. Over the past few decades, burgeoning academic literature argues or assumes that past societies’ responses to environmental impacts can provide valuable lessons to guide adaptation to contemporary environmental changes. The ancient Maya civilization, whose decline is often linked with historically unprecedented droughts, constitutes a major case study for extracting such lessons. To analyze these lessons, we conduct a meta-analysis of the discourse of learning from past Maya-environment interactions. We demonstrate that although studies often refer to learning from the Maya explicitly, the learning is primarily declarative and discursive rather than substantive, and lessons are often vague, misguided, or inapplicable. Only a few articles employ research designs conducive to learning from the past, and only a few reflect on the process of, or the problems associated with, learning from the past. On the other hand, many articles are content with reaching ‘inspirational’ lessons, calling, for example, for increasing resilience, while only a fifth of the papers drew more specific lessons that offer somewhat concrete recommendations and courses of action. Many studies also claimed that their findings are applicable to present-day societies far outside the core regions of Maya habitation, ignoring pertinent social and geographical differences. Although the paper does not preclude the theoretical possibility of learning from the past, it argues that such learning must undergo significant changes to achieve robustness and relevance for the present. This would also require a more open discussion between scholars of the past and adaptation practitioners.

The Role of Memory in LLMs: Persistent Context for Smarter Conversations

Memory in LLMs has given way to more logical and sensible interactions between the system and the user. This is different from other models that are session bound, such that the responses to any one query are not related to past and future interactions with the same user, but memory-enabled LLMs retain information across sessions and continually update interactions with the person they are communicating with. The role of permanent memory in LLMs is considered in this work, provided through the analysis of the role of memory mechanisms in maintaining conversation flows, improving user interaction, and supporting practical applications in various industries, including customer service, healthcare, and education. Discussing how the idea and architectures of memory correspond to storage and retrieval procedures and the management of memory in LLMs this paper outlines the opportunities and challenges for AI systems that want to include contextual intelligence but also remain ethical. The synthesis of important concepts underlines the promising prospects of memory-augmented models in improving the communication with users and points to the imperatively important aspect of controlling the memory process at the design stage of LLMs. We also offer recommendations for privacy and ethical concerns that should be avoided in the case of future AI memory advancements in an effort to pursue sustainable technological progress while also incorporating user-oriented values into the process.

I have seen this before: imprinting experiences and Bank CEO risk-taking in times of crisis

Abstract Understanding the causes and consequences of corporate risk-taking has remained a crucial topic for organizational scholars. Using the case of U.S. banks and one dimension of their risk-taking behavior around the 2008 financial crisis, we offer a theory of how the diverse experiences of corporate leaders can shape their risk-taking behavior. Building on the imprinting literature, we theorize how different types of experiences that bank CEOs had in the past interact to shape current risk-taking behavior, resulting in risk moderation under crisis conditions. We focus on two imprinting experiences with particular relevance for bank CEOs’ risk-taking behavior—MBA education and past crisis experience. We argue that the latter played a pronounced role during the crisis because of greater imprint-environment fit. Our analysis using data from 170 large banks between 2001 and 2019 shows that bank CEOs’ firsthand experience of a prior banking crisis not only directly tempered bank risk-taking but also did so indirectly by limiting the risk-taking tendencies of CEOs with an MBA, particularly during the crisis period. Our study contributes to the sociological literature about organizational risk-taking by emphasizing the crucial role of organizational leaders’ biographies and exploring how earlier institutional conditions shape their risk-taking behavior later.

DHKFN

Knowledge tracing is effective in modeling learners' knowledge levels to predict future answering situations based on their past learning history and interaction processes. However, current methods often overlook the impact of knowledge point correlations on answer prediction. This paper design a model based on deep hierarchical knowledge fusion networks (DHKFN). It utilizes statistical approaches to construct correlation matrices to capture the correlations between knowledge points. Subsequently, it constructs a topology graph structure of questions and knowledge points and utilizes a multi-head attention network to learn student interaction information from multiple perspectives, effectively constructing adjacency matrices. Finally, it uses GCN to learn the interaction information representation between deep-level knowledge points from dynamically constructed a topology graph structures. Experimental results on three large public datasets show that DHKFN effectively considers the influence of correlations between different knowledge points, thus showing promising effectiveness.

Spacetime symmetry breaking on nongeodesic leaves and a new form of matter

We examine the permanent damage caused by the historical breakdown of full diffeomorphism invariance induced by a foliation. We focus on the case where the foliation is allowed to be nongeodesic after the interactions with the foliation switch off. Gravity and other forms of matter recover full diffeomorphism invariance only at the expense of introducing a new matterlike component, carrying the nonvanishing Hamiltonian (and momentum, as it turns out) left over from the violating past interactions. This matter form must be stress-free in the preferred frame; this is the only way a matter action can mimic the evolution of the leftover Hamiltonian (and momentum) driven by the Dirac hypersurface deformation algebra. Hence, if the preferred frame is nongeodesic, the equivalent matter component must have energy and a momentum current in this frame, but still no spatial stresses: an unusual form of “matter.” It is equivalent to a fluid with anisotropic stress in some regimes, reducing to dust in others, or even displaying completely new features in extreme situations. Its stress energy tensor is conserved. We provide two examples based on accelerated frames: Rindler space-time and the canonical Schwarzchild frame. Published by the American Physical Society 2024

An investigation of toddler risky play informed by the voices of parents and teachers of toddlers in one school setting

ABSTRACT This qualitative study explored the perceptions of teachers and parents within the context of toddlerhood risky play at a university laboratory school. Sociocultural and bioecological theories informed the study and the developmental niche framework guided the data analysis. Video-stimulated recall interviews were analyzed using the constant comparative method. Findings include: (1) parents’ and teachers’ perceptions of risky play are comprised of (a) their psychology as the caregiver, (b) knowing the child, and (c) cultural influences, (2) parent and teacher descriptions of risky play include that it happens outdoors, away from adults, and encompasses developmental benefits, and (3) the development of perceptions of risky play is situated and dynamic. This study contributes important knowledge to the field of early childhood education by offering a new perspective regarding the definition of risky play revealing how parents’ and teachers’ perceptions of risky play are situated in their own past experiences, knowledges, and interactions.

Macroscale Roughness Reveals the Complex History of Asteroids Didymos and Dimorphos

Morphological mapping is a fundamental step in studying the processes that shaped an asteroid surface. However, it is challenging and often requires multiple independent assessments by trained experts. Here we present fast methods to detect and characterize meaningful terrains from the topographic roughness: entropy of information, and local mean surface orientation. We apply our techniques to Didymos and Dimorphos, the target asteroids of NASA's Double Asteroid Redirection Test mission—the first attempt to deflect an asteroid. Our methods reliably identify morphological units at multiple scales. The comparative study reveals various terrain types, signatures of processes that transformed Didymos and Dimorphos. Didymos shows the most heterogeneity and morphology that indicate recent resurfacing events. Dimorphos is comparatively rougher than Didymos, which may result from the formation process of the binary pair and past interaction between the two bodies. Our methods can be readily applied to other bodies and data sets.

Beyond hinges and spires: A critical examination of archaeomalacological quantification methodologies using coral reef molluscan assemblages from Jiigurru (Lizard Island Group), northern Great Barrier Reef

Quantification, as one of the pillars of the zooarchaeological subdiscipline, is an invaluable component of the toolkit researchers use to study past people-animal interactions. Despite being the subject of rigorous (zoo)archaeological debate, the calculation of Minimum Number of Individuals (MNI) values remains one of the main methods of quantifying the relative abundance of taxa within faunal assemblages. Choosing the appropriate quantification protocol to calculate the MNI of archaeological invertebrate assemblages can be challenging due to regional taxonomic considerations and the myriad of quantification methodologies and frameworks available in the global archaeomalacological literature. In an Australian context, methodologies for quantifying coral reef molluscan assemblages have not been explicitly evaluated. Using archaeological molluscan assemblages from two midden sites (Freshwater Bay Midden and Mangrove Beach Headland Midden) on the northern Great Barrier Reef island group of Jiigurru (the Lizard Island Group), we critically examine two commonly adopted archaeomalacological quantification methodologies: the NRE MNI and tMNI protocols. The NRE MNI methodology uses one to two non-repetitive elements (NREs) of molluscs, whilst the tMNI protocol includes a wider range of elements akin to vertebrate MNI quantification methodologies. Through a comparison of taxa abundances and statistical analyses, results show that the tMNI protocol, with some modification, is best suited for the Jiigurru assemblages. Higher MNI values and an increased assemblage diversity, evenness, and richness were recorded for the molluscan assemblages at both midden sites when the tMNI protocol was applied. This study foregrounds the importance of data transparency when reporting quantification protocols and outcomes to ensure the highest degree of data quality, replicability, and usability.

Bronze Age cheese reveals human-Lactobacillus interactions over evolutionary history

Despite the long history of consumption of fermented dairy, little is known about how the fermented microbes were utilized and evolved over human history. Here, by retrieving ancient DNA of Bronze Age kefir cheese (∼3,500 years ago) from the Xiaohe cemetery, we explored past human-microbial interactions. Although it was previously suggested that kefir was spread from the Northern Caucasus to Europe and other regions, we found an additional spreading route of kefir from Xinjiang to inland East Asia. Over evolutionary history, the East Asian strains gained multiple gene clusters with defensive roles against environmental stressors, which can be a result of the adaptation of Lactobacillus strains to various environmental niches and humanselection. Overall, our results highlight the role of past human activities in shaping the evolution of human-related microbes, and such insights can, in turn, provide a better understanding of past human behaviors.

Egalitarian preferences in young children depend on the genders of the interacting partners

In decisions between equal and unequal resource distributions, women are often believed to be more prosocial than men. Previous research showed that fairness attitudes develop in childhood, but their—possibly gendered, developmental trajectory remains unclear. We hypothesised that gender-related fairness attitudes might depend not only on the gender of the Allocator, but also on that of the Recipient. To examine this, we tested 332 three to 8-year-old children in a paired resource allocation task, with both boys and girls acting as Allocators and Recipients. We indeed found gender-related effects: girls more than boys aimed to reduce advantageous inequity, and Allocators of both genders were more averse against male Recipients being better off. Notably, older girls exhibited an envy bias, i.e., they tolerated disadvantageous inequity more when the resource allocation was in favour of other girls than when it favoured boys. We also observed a gender-related spite gap in boys aged 7-8: unlike girls, boys treated other boys with spite, i.e., they valued unfair distributions in their own favour over equal outcomes, especially if rejecting advantageous inequity was costly. This pattern hints at contextualised gender-related fairness preferences that evolve with age that could depend on same- and cross-gender past interaction experiences.

Predicting anxiety treatment outcome in community mental health services using linked health administrative data

Anxiety disorders is ranked as the most common class of mental illness disorders globally, affecting hundreds of millions of people and significantly impacting daily life. Developing reliable predictive models for anxiety treatment outcomes holds immense potential to help guide the development of personalised care, optimise resource allocation and improve patient outcomes. This research investigates whether community mental health treatment for anxiety disorder is associated with reliable changes in Kessler psychological distress scale (K10) scores and whether pre-treatment K10 scores and past health service interactions can accurately predict reliable change (improvement). The K10 assessment was administered to 46,938 public patients in a community setting within the Western Australia dataset in 2005–2022; of whom 3794 in 4067 episodes of care were reassessed at least twice for anxiety disorders, obsessive–compulsive disorder, or reaction to severe stress and adjustment disorders (ICD-10 codes F40–F43). Reliable change on the K10 was calculated and used with the post-treatment score as the outcome variables. Machine learning models were developed using features from a large health service administrative linked dataset that includes the pre-treatment K10 assessment as well as community mental health episodes of care, emergency department presentations, and inpatient admissions for prediction. The classification model achieved an area under the receiver operating characteristic curve of 0.76 as well as an F1 score, precision and recall of 0.69, and the regression model achieved an R2 of 0.37 with mean absolute error of 5.58 on the test dataset. While the prediction models achieved moderate performance, they also underscore the necessity for regular patient monitoring and the collection of more clinically relevant and contextual patient data to further improve prediction of treatment outcomes.

Local-time formula for dissipation in solid ionic electrolytes

When ions move through solids, they interact with the solid's constituent atoms and cause them to vibrate around their equilibrium points. This vibration, in turn, modifies the potential landscape through which the mobile ions travel. Because the present-time potential depends on past interactions, the coupling is inherently nonlocal in time, making its numerical and analytical treatment challenging. For sufficiently slow-moving ions, we linearize the phonon spectrum to show that these nonlocal effects can be ignored, giving rise to a draglike force. Unlike the more familiar drag coefficient in liquids, the drag takes on a matrix form due to the crystalline structure of the framework. We numerically simulate trajectories and dissipation rates using both the time-local and nonlocal formulas to validate our simplification. The time-local formula dramatically reduces the computational cost of calculating the motion of a mobile particle through a crystalline framework and clearly connects the properties of the material to the drag experienced by the particle. Published by the American Physical Society 2024

IDC-CDR: Cross-domain Recommendation based on Intent Disentanglement and Contrast Learning

Using the user’s past activity across different domains, the cross-domain recommendation (CDR) predicts the items that users are likely to click. Most recent studies on CDR model user interests at the item level. However because items in other domains are inherently heterogeneous, direct modeling of past interactions from other domains to augment user representation in the target domain may limit the effectiveness of recommendation. Thus, in order to enhance the performance of cross-domain recommendation, we present a model called Cross-domain Recommendation based on Intent Disentanglement and Contrast Learning (IDC-CDR) that performs contrastive learning at the intent level between domains and disentangles user interaction intents in various domains. Initially, user–item interaction graphs were created for both single-domain and cross-domain scenarios. Then, by modeling the intention distribution of each user–item interaction, the interaction intention graph and its representation were updated repeatedly. The comprehensive local intent is then obtained by fusing the local domain intents of the source domain and the target domain using the attention technique. In order to enhance representation learning and knowledge transfer, we ultimately develop a cross-domain intention contrastive learning method. Using three pairs of cross-domain scenarios from Amazon and the KuaiRand dataset, we carry out comprehensive experiments. The experimental findings demonstrate that the recommendation performance can be greatly enhanced by IDC-CDR, with an average improvement of 20.62% and 25.32% for HR and NDCG metrics, respectively.

An efficient state-aware Coarse-Fine-Grained model for Knowledge Tracing

Knowledge tracing (KT) is the problem of predicting students’ future performance based on their historical learning records, with primary emphasis on capturing their evolving knowledge states. This predictive capacity enables online education platforms to recommend personalized learning resources and guidance to students who require further attention. By leveraging deep learning techniques, KT models have demonstrated impressive experimental performance in capturing and analyzing the complex interactions between students and exercises. However, most approaches tend to prioritize predictive performance rather than a detailed consideration of how students’ knowledge states evolve over time, resulting in unstable and unexplainable outcomes. To address these limitations, in this paper, we propose an efficient state-aware Coarse-Fine-Grained model for Knowledge Tracing (CFGKT), which aims to facilitate deep tracing of students’ intrinsic knowledge states in their learning processes. Specifically, CFGKT introduces three key innovations: (1) a coarse-grained block that diagnoses students’ proficiency in specific concepts based on their historical performance; (2) a fine-grained block that progressively measures and accumulates students’ experience acquisition from past interactions; and (3) a novel state-fusion attention that selectively integrates both exclusive proficiency and cumulative experience to estimate students’ knowledge states for performance prediction. Numerous experiments on four real-world datasets demonstrate that CFGKT exhibits superior prediction accuracy and interpretability compared to existing KT methods. The code and data are publicly available at https://github.com/hzLu0/CFGKT.

Sequential query prediction based on multi-armed bandits with ensemble of transformer experts and immediate feedback

We study the problem of predicting the next query to be recommended in interactive data exploratory analysis to guide users to correct content. Current query prediction approaches are based on sequence-to-sequence learning, exploiting past interaction data. However, due to the resource-hungry training process, such approaches fail to adapt to immediate user feedback. Immediate feedback is essential and considered as a signal of the user’s intent. We contribute with a novel query prediction ensemble mechanism, which adapts to immediate feedback relying on multi-armed bandits framework. Our mechanism, an extension to the popular Exp3 algorithm, augments Transformer-based language models for query predictions by combining predictions from experts, thus dynamically building a candidate set during exploration. Immediate feedback is leveraged to choose the appropriate prediction in a probabilistic fashion. We provide comprehensive large-scale experimental and comparative assessment using a popular online literature discovery service, which showcases that our mechanism (i) improves the per-round regret substantially against state-of-the-art Transformer-based models and (ii) shows the superiority of causal language modelling over masked language modelling for query recommendations.

Skip-Gram and Transformer Model for Session-Based Recommendation

Session-based recommendation uses past clicks and interaction sequences from anonymous users to predict the next item most likely to be clicked. Predicting the user’s subsequent behavior in online transactions becomes a problem mainly due to the lack of user information and limited behavioral information. Existing methods, such as recurrent neural network (RNN)-based models that model user’s past behavior sequences and graph neural network (GNN)-based models that capture potential relationships between items, miss different time intervals in the past behavior sequence and can only capture certain types of user interest patterns due to the characteristics of neural networks. Graphic models created to improve the current session reduce the model’s success due to the addition of irrelevant items. Moreover, attention mechanisms in recent approaches have been insufficient due to weak representations of users and products. In this study, we propose a model based on the combination of skip-gram and transformer (SkipGT) to solve the above-mentioned drawbacks in session-based recommendation systems. In the proposed method, skip-gram both captures chained user interest in the session thread through item-specific subreddits and learns complex interaction information between items. The proposed method captures short-term and long-term preference representations to predict the next click with the help of a transformer. The transformer in our proposed model overcomes many limitations in turn-based models and models longer contextual connections between items more effectively. In our proposed model, by giving the transformer trained item embeddings from the skip-gram model as input, the transformer has better performance because it does not learn item representations from scratch. By conducting extensive experiments with three real-world datasets, we confirm that SkipGT significantly outperforms state-of-the-art solutions with an average MRR score of 5.58%.

How is feedback perceived by Brazilian students and faculty from a nursing school?

AimThis study aims to explore the perceptions of feedback among undergraduate students and faculty members at a Brazilian private nursing school. BackgroundFeedback plays a crucial role in the socioemotional development of learners, with its interpretation varying across different sociocultural contexts. Student evaluations frequently express dissatisfaction regarding both the quality and quantity of feedback received. Conversely, delivering feedback poses a challenge for faculty, requiring the establishment of an empathetic connection that fosters trust and credibility. Brazil, being a developing country characterized by social disparities and economic challenges, presents a unique backdrop for examining feedback dynamics. DesignQualitative research, employing Inductive Content Analysis, was used to understand feedback perceptions in Brazilian nursing education. Symbolic interactionism was adopted as methodological framework and guided data interpretation. MethodsWe carried out five virtual focus groups composed of a group of teachers (n=5) and four of students (n=34). Semi-structured interviews guided data collection. The recorded sessions were subsequently analyzed to identify key themes and codes. Symbolic interactionism was employed as a framework to derive meaning from qualitative data. ResultsContent analysis generated two categories that reveal the perception of teachers and students in the feedback process. The first, called “Feedback in Education: Sociocultural Influences for Students and Teachers”, expresses the beliefs and interpretations of students and teachers within the shared feedback environment. The second called “Challenging resonance, transformative construction: Navigating the dualities of feedback for teachers and students”, which elucidated how relational dynamics shape behaviors and attitudes, promoting the development of social skills and learning. Faculty's previous feedback experiences significantly influence their self-perception and behavior with students. As a result of the resonance of these past interactions, we recognize that the teaching self also plays a crucial role in the quality and perception of feedback. Furthermore, students construct social reality with similar beliefs and values, they believe in the learning potential generated by feedback. Our findings also corroborate that perceptions of feedback are deeply influenced by the sociocultural context and the narratives corroborate previous findings indicating that, in Brazil, honest feedback can be implicitly perceived as criticism rather than an opportunity for growth. ConclusionsFaculty members often draw on their past experiences when providing feedback, highlighting the adaptive nature of feedback interactions. Additionally, the feedback process is consistently influenced by the commitment to maintaining positive relationships with students. Students recognize the constructive dimension of feedback as a valuable tool for learning and personal growth.

An international, open-access dataset of dental wear patterns and associated broad age classes in archaeological cattle mandibles

Zooarchaeologists investigate past interactions between animals, humans, and their environments by analyzing the remains of archaeological fauna. Age-at-death distributions are fundamental to faunal analysis and are often estimated by comparing exposed dentine patterns to standardized tooth wear stages that have been associated with relative age classes. We present Bubona, an international dataset of dental wear patterns and associated broad age classes in archaeological cattle mandibles. Our open-access dataset of 1460 data entries from nine counties is being used to create tooth-type specific reference tables of probable age class attribution for cattle mandibles lacking complete dentition. Bubona is a valuable resource for the innovation of new systems of age estimation for cattle and it is the creators hope that researchers will continue to both help expand the dataset by contributing their own data, as well as utilize the data to refine and innovate age-at-death estimation methods.

Contrastive cross-domain sequential recommendation via emphasized intention features

The objective of cross-domain sequential recommendation is to forecast upcoming interactions by leveraging past interactions across diverse domains. Most methods aim to utilize single-domain and cross-domain information as much as possible for personalized preference extraction and effective integration. However, on one hand, most models ignore that cross-domain information is composed of multiple single-domains when generating representations. They still treat cross-domain information the same way as single-domain information, resulting in noisy representation generation. Only by imposing certain constraints on cross-domain information during representation generation can subsequent models minimize interference when considering user preferences. On the other hand, some methods neglect the joint consideration of users’ long-term and short-term preferences and reduce the weight of cross-domain user preferences to minimize noise interference. To better consider the mutual promotion of cross-domain and single-domains factors, we propose a novel model (C2DREIF) that utilizes Gaussian graph encoders to handle information, effectively constraining the correlation of information and capturing useful contextual information more accurately. It also employs a Top-down transformer to accurately extract user intents within each domain, taking into account the user’s long-term and short-term preferences. Additionally, entropy regularized is applied to enhance contrastive learning and mitigate the impact of randomness caused by negative sample composition.

A comparison between different models of delivering maternal cash transfers in Myanmar.

As part of a randomized controlled trial conducted in Myanmar between 2016 and 2019, we explore the performance of a maternal cash transfer program across villages assigned to different models of delivery (by government health workers vs loan agents of a non-governmental organization) and identify key factors of success. Measures include enrolment inclusion and exclusion errors, failures in payment delivery to enrolled beneficiaries (whether beneficiaries received any transfer, fraction of benefits received and whether there were delays and underpayment of benefit amounts) and whether beneficiaries remained in the program beyond eligibility. We find that women in villages where government health workers delivered cash transfers received on average two additional monthly transfers, were 19.7% more likely to receive payments on time and in-full and were 14.6% less likely to stay in the program beyond eligibility. With respect to the primary health objective of the program-child nutrition-we find that children whose mother received cash by government health workers were less likely to be chronically malnourished compared to those whose mother received cash by loan agents. Overall, the delivery of cash transfers to mothers of young children by government health workers outperforms the delivery by loan agents in rural Myanmar. Qualitative evidence suggests two key factors of success: (1) trusted presence and past interactions with targeted beneficiaries and complementarities between government health workers' expertise and the program; and (2) performance incentives based on specific health objectives along with top-down monitoring. We cannot exclude that other incentives or intrinsic motivation also played a role.