43569 research outputs found
Sort by
Measurement of teacher’s orchestration load: a framework and a case study on tool flexibility
No full textTeacher orchestration in Computer-Supported Collaborative Learning (CSCL) environments demands managing multiple tasks across different social levels, often under tight constraints, leading to an increased orchestration load. This load represents the cognitive and physical effort teachers invest in real-time coordination of learning activities, which remains underexplored, particularly how the flexibility of orchestration tools influences this burden. In response to this gap, we propose a comprehensive framework for tracking and characterizing orchestration load, focusing on the intensity and dynamics of teachers’ actions while implementing CSCL scripts. The framework integrates multimodal data sources, including observable orchestration actions, physiological metrics, and self-reported insights, to comprehensively analyze the orchestration load. We illustrate the applicability of the framework through a case study comparing two CSCL orchestration tools: PyramidApp, which offers pre-configured, structured enactment of the Pyramid collaborative learning flow pattern, and EthicApp, a more flexible tool allowing real-time design adjustments for a variety of instructional designs. Our findings reveal that PyramidApp facilitates a streamlined orchestration process with lower intensity and reduced teacher workload. At the same time, EthicApp’s high flexibility increases orchestration intensity, requiring more cognitive effort from teachers during real-time phase configuration. These results underscore the importance of balancing flexibility with usability in orchestration tool design, as overly flexible environments may overwhelm teachers, especially during high cognitive load scenarios. This study contributes a methodological framework for evaluating orchestration tools and highlights key design considerations for reducing orchestration load in CSCL environments.This work was supported in part by Ministerio de Ciencia, Innovación y Universidades (MICIU)/Agencia Estatal de Investigación (AEI)/10.13039/501100011033 under Grant PID2020-112584RB-C33, Grant PID2023-146692OB-C33, and Grant CEX2021-001195-M; in part by Support to Research Groups (SGR) under Grant 00930; and in part by the Annual Internal competition for Research support funds, ‘‘RESEARCH PROJECTS COMPETITION - FEN 2020’’, Faculty of Economics and Business (FEN), University of Chile. The work of Davinia Hernandez-Leo was supported by Institución Catalana de Investigación y Estudios Avanzados (ICREA) Academia
Nonhypermutator cancers access driver mutations through reversals in germline mutational bias
No full textCancer is an evolutionary disease driven by mutations in asexually reproducing somatic cells. In asexual microbes, bias reversals in the mutation spectrum can speed adaptation by increasing access to previously undersampled beneficial mutations. By analyzing tumors from 20 tissues, along with normal tissue and the germline, we demonstrate this effect in cancer. Nonhypermutated tumors reverse the germline mutation bias and have consistent spectra across tissues. These spectra changes carry the signature of hypoxia, and they facilitate positive selection in cancer genes. Hypermutated and nonhypermutated tumors thus acquire driver mutations differently: hypermutated tumors by higher mutation rates and nonhypermutated tumors by changing the mutation spectrum to reverse the germline mutation bias.This work was supported by the Natural Sciences and Engineering Research Council of Canada grant RGPIN-2019-06294, by the National Institute of General Medical Sciences of the National Institutes of Health through grants R01GM127348 and R35GM149235, and by the Spanish Ministry of Science and Innovation through the Centro de Excelencia Severo Ochoa (CEX2020-001049-S, MCIN/AEI /10.13039/501100011033), the Generalitat de Catalunya through the CERCA programme, and the European Union’s H2020 research and innovation program under Marie Sklodowska-Curie grant agreement No.754422
Compositional brain scores capture Alzheimer's disease-specific structural brain patterns along the disease continuum
No full textIntroduction: Traditional multivariate methods for neuroimaging studies overlook the interdependent relationship between brain features. This study addresses this gap by analyzing relative brain volumetric patterns to capture how Alzheimer's disease (AD) and genetics influence brain structure along the disease continuum. Methods: This study analyzed data from participants across the AD continuum from the Alzheimer's and Families (ALFA) and Alzheimer's Disease Neuroimaging Initiative (ADNI) studies. Compositional data analysis (CoDA) was exploited to examine relative brain volumetric variations that (1) were linked to different AD stages compared to cognitively unimpaired amyloid-β-negative (CU A-) individuals and (2) varied by AD genetic risk. Results: Disease stage-specific compositional brain scores were identified, differentiating CU A- individuals from those in more advanced stages. Genetic risk-stratified models revealed a broader genetic landscape affecting brain morphology in AD, beyond the well-known apolipoprotein E ε4 allele. Discussion: CoDA emerges as an alternative multivariate framework to deepen understanding of AD-related structural changes and support targeted interventions for those at higher genetic risk. Highlights: Compositional data analysis (CoDA) revealed the relative variation of brain region volumes, captured in compositional brain scores, capable of discerning between cognitively unimpaired amyloid-β-negative individuals and subjects within other disease-stage groups along the Alzheimer's disease (AD) continuum. CoDA also uncovered the genetic vulnerability of specific brain regions at each stage of the disease along the continuum. CoDA is capable of integrating magnetic resonance imaging data from two different cohorts without stringent requirements for harmonization. This translates as an advantage, compared to traditional methods, and strengthens the reliability of cross-study comparisons by standardizing the data despite different labeling agreements, facilitating collaborative and large-scale research. The algorithm is sensitive to AD-specific effects, as the main compositional brain scores display little overlap with the age-specific compositional brain score. CoDA provides a more accurate analysis of brain imaging data addressing its compositional nature, which can influence the development of targeted approaches, opening new avenues for enhancing brain health.The research leading to these results has received funding from “la Caixa” Foundation (ID 100010434), under agreement LCF/PR/GN17/50300004, the Health Department of the Catalan Government (Health Research and Innovation Strategic Plan (PERIS) 2016-2020 grant# SLT002/16/00201), and the Alzheimer's Association and an international anonymous charity foundation through the TriBEKa Imaging Platform project (TriBEKa-17-519007). Additional support has been received from the Universities and Research Secretariat, Ministry of Business and Knowledge of the Catalan Government under the grant no. 2021 SGR 00913. All CRG authors acknowledge the support of the Spanish Ministry of Science, Innovation, and Universities to the EMBL partnership, the Centro de Excelencia Severo Ochoa, and the CERCA Programme/Generalitat de Catalunya. N.V.-T. was supported by the Spanish Ministry of Science and Innovation—State Research Agency (IJC2020-043216-I/MCIN/AEI/10.13039/501100011033) and the European Union «NextGenerationEU»/PRTR and currently receives funding from the Spanish Research Agency MICIU/AEI/10.13039/501100011033 (grant RYC2022-038136-I cofunded by the European Union FSE+ and grant PID2022-143106OA-I00 cofunded by the European Union FEDER). In addition, N.V.-T. is supported by the William H. Gates Sr. Fellowship from the Alzheimer's Disease Data Initiative. Data partially used in preparation of this article were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database (adni.loni.usc.edu). As such, the investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report
Early life exposure to fine particulate matter and fine motor function, attentional function, and working memory among Spanish school-aged children
No full textBackground: Evidence of the association between fine particulate matter (PM2.5) exposure and child neuropsychological function is equivocal. We examined early life PM2.5 exposure in relation to fine motor function, attention, and working memory in early childhood. Methods: We used data from the Spanish INfancia y Medio Ambiente Project, 2003-2008. Exposure to PM2.5 (μg/m3) was assessed using spatiotemporal land-use random forest models and assigned based on residential address histories. Around age six, children completed the finger tapping test, attentional network test (ANT), and n-back task to evaluate fine motor speed, attention, and working memory, respectively. A total of 1,310 children had data from at least one neuropsychological assessment. General linear models were applied to assess associations between average prenatal and postnatal PM2.5 with each outcome. Distributed lag nonlinear models were used to explore refined periods of susceptibility to PM2.5. We reported β estimates and 99% credible intervals (CrI) representing the change in each outcome per 5-μg/m3 increase in PM2.5. Results: Prenatal PM2.5 exposure was associated with decreased mean hit reaction time (HRT) (β = -21.82; 99% CrI = -64.1, 20.4) and HRT-standard error (β = -9.7; 99% CrI = -30.3, 10.9) on the ANT but estimates were imprecise. Postnatal PM2.5 was associated with reduced mean HRT on the n-back task (β = -39.4; 99% CrI = -115.1, 26.3). We observed sensitive periods of exposure in the postnatal period associated with both better and worse performance on the finger-tapping test and ANT. Conclusions: We found limited evidence to support an association between PM2.5 exposure and fine motor function, attentional function, or working memory in school-aged children.The results reported herein correspond to the specific aims of grant R01ES028842 to K.W.W. from the National Institutes of Health/National Institute of Environmental Health Sciences (NIH/NIEHS). This work was also supported by grants Red INMA G03/176, CB06/02/004; ISCIII-FEDER: PI03/1615, PI04/1509, PI04/1112, PI04/1931, PI04/2018, PI05/1079, PI05/1052, PI06/0867, PI06/1213, PI07/0314, PI09/02311, PI09/02647, PI11/01007, PI11/02591, PI11/02038, PI13/1944, PI13/2032, PI13/02429, PI14/00891, PI14/01687, PI16/1288, PI16/00118, PI17/00663, PI18/00909, PI18/01142, and PI18/01237; Miguel Servet-FEDER CP11/00178, CP15/00025, CPII16/00051, CPII18/00018, and CP16/00128 from Instituto de Salud Carlos III, grant 1999SGR 00241from Generalitat de Catalunya-CIRIT, grant FP7-ENV-2011 cod 282957 and HEALTH.2010.2.4.5-1 from the EU Commission, Assistance Award No. R-82811201 from the Health Effects Institute, grant UGP-15-230, UGP-15-244, and UGP-15-249 from Generalitat Valenciana: FISABIO, grant 2005111093 from Alicia Koplowitz Foundation 2017, Department of Health of the Basque Government, grant DFG06/002 from the Provincial Government of Gipuzkoa, and annual agreements with the municipalities of the study area (Zumarraga, Urretxu, Legazpi, Azkoitia y Azpeitia y Beasain), and Margarita Salas Grant MS21-125 co-funded by European Union- Next Generation EU and Ministerio de Universidades. We also acknowledge support from the Spanish Ministry of Science and Innovation and the State Research Agency through the “Centro de Excelencia Severo Ochoa 2019-2023” Program (CEX2018-000806-S), and support from the Generalitat de Catalunya through the CERCA Program. K.W.W. and E.S. were partially supported by the P30 Environmental Health Sciences Core Center grant P30ES030285 from the NIH/NIEHS and by funding from the NIH/National Institute on Minority Health and Health Disparities (NIMHD) under Award Number P50MD015496
La construcció mediàtica del món casteller a Catalunya: anàlisi dels mitjans locals, generalistes i especialitzats
No full textTutor: Xavier Ramon VegasTreball de fi de grau en Periodisme. Curs 2024-2025Els castells s’han consolidat al llarg de la seva història com una de les expressions més atractives de la cultura catalana gràcies a la seva espectacularitat visual i els seus valors. Amb l’evolució d’aquesta activitat van anar apareixent espais en el sistema mediàtic català que tractaven l’actualitat informativa del món casteller. Aquests espais i els mitjans han anat evolucionant amb l’arribada de la digitalització al periodisme que n’ha canviat la cobertura amb l’aparició de nous formats comunicatius i suports digitals. Aquest treball analitza com es comunica actualment l’activitat castellera als mitjans de comunicació de Catalunya, posant el focus en el tractament informatiu de les diades castelleres. L’estudi persegueix dos objectius principals: analitzar i comparar la cobertura que en fan mitjans locals, generalistes i especialitzats; a més de descriure com es construeixen les peces informatives, coneixent els perfils professionals implicats i els processos de treball. Per fer-ho, s’ha combinat una anàlisi quantitativa de 360 peces informatives (procedents de premsa escrita i digital, televisió i ràdio) amb una anàlisi qualitativa basada en entrevistes en profunditat a cinc periodistes especialitzats en el món casteller. Aquesta recerca respon a la necessitat d’investigar un àmbit poc explorat en la bibliografia acadèmica actual que estudia l’activitat castellera principalment des de camps d’estudi com l’antropologia, l’etnografia i la sociologia.Throughout their history, human towers have established themselves as one of the most attractive expressions of Catalan culture, thanks to their visual spectacularity and the values they represent. As this activity evolved, spaces emerged in the Catalan media system that covered news from the “casteller” world. These spaces and the media have evolved alongside the digitalisation of journalism, which has transformed the way it covers events with the emergence of new communication formats and digital media. This study examines how human tower activities are currently reported in Catalan media, with a focus on the coverage of human tower days. The study pursues two main objectives: analysing and comparing the coverage provided by local, generalist and specialised media, and describing how information pieces are constructed, identifying the professional profiles involved and the work processes. To achieve this, a quantitative analysis of 360 news items from the 2 written and digital press, television and radio has been combined with a qualitative analysis based on in-depth interviews with five journalists specialising in the “casteller” world. This research addresses the lack of academic literature on human tower activity, which is mainly studied from the perspectives of anthropology, ethnography and sociology
Metastable dynamics emerge from local excitatory–inhibitory homeostasis in the cortex at rest
No full textThe dynamics of the human cortex are highly metastable, driving the spontaneous exploration of network states. This metastability depends on circuit-level edge-of-bifurcation dynamics, which emerge from firing-rate control through multiple mechanisms of excitatory–inhibitory (E–I) homeostasis. However, it is unclear how these contribute to the metastability of cortical networks. We propose that individual mechanisms of the E–I homeostasis contribute uniquely to the emergence of resting-state dynamics and test this hypothesis in a large-scale model of the human cortex. We show that empirical connectivity and dynamics can only be reproduced when accounting for multiple mechanisms of the E–I homeostasis. More specifically, while the homeostasis of excitation and inhibition enhances metastability, the regulation of intrinsic excitability ensures moderate synchrony, maximizing functional complexity. Furthermore, the modulation bifurcation modulation by the homeostasis of excitation and intrinsic excitability compensates for strong input fluctuations in connector hubs. Importantly, this only occurs in models accounting for local gamma oscillations, suggesting a relationship between E–I balance, gamma rhythms, and metastable dynamics. Altogether, our results show that cortical networks self-organize toward maximal metastability through the multifactor homeostasis of E–I balance. Therefore, the benefits of combining multiple homeostatic mechanisms transcend the circuit level, supporting the metastable dynamics of large-scale cortical networks
Annotating the microbial dark matter with HiFi-NN
No full textThe accurate computational annotation of protein sequences with enzymatic function remains a fundamental challenge in bioinformatics. Here, we present HiFi-NN (Hierarchically-Finetuned Nearest Neighbor search) which annotates protein sequences to the 4th level of Enzyme Commission (EC) number with greater precision and recall than state-of-the-art deep learning methods. Furthermore, we show that this method can correctly identify the EC number of a given sequence to lower identities than BLASTp. We show that performance can be improved by increasing the diversity of the lookup set in both sequence space and the environment the sequence has been sampled from. We proceed to show that we can correct specific mis-annotations in the BRENDA enzymes database reproducing results found by others. Finally, we use HiFi-NN to annotate functional dark-matter protein sequences from NMPFamDB. Our findings pave the way for more accurate functional annotation in silico, especially for proteins from distant sequence space
Exploring the integration of large language models for automatic emotion labeling in speech
No full textTreball fi de màster de: Master in Intelligent Interactive SystemsSupervisor: Prof. María Inés Torres BarañanoIn this work, we present a comprehensive comparison of methodologies for speech emotion recognition (SER), with a focus on evaluating the effectiveness of large language models (LLMs) in this domain. Our study is structured into three parts. First, we extract audio embeddings using models such as WavLM, HuBERT, and Dasheng, and use classical machine learning classifier-Support Vector Machine (SVM) and
Multilayer Perceptron (MLP) for emotion prediction. These approach serves as a baseline for comparison. Second, we investigate the capacity of LLMs like GPT-4o, Qwen2-Audio, and Amazon Nova Sonic to analyze audio features, including speaker attributes such as gender, thereby extending their application beyond traditional natural language processing. Third, we explore a more integrated approach that
directly inputs raw audio into LLM for audio processing, such as Qwen2-Audio7B-Instruct, for end-to-end emotion classification, without the need for traditional signal-processing-based feature extraction. We evaluate and compare the performance of these methodologies based on various metrics, such as accuracy, precision, recall, and F1-score. A key aspect of this study is the primary focus on the results obtained from LLM-based models. Our results reveal several key insights: (1) data distribution significantly affects classifier performance; (2) different audio embeddings shows different results even with the same classifier and dataset; and (3) despite their capability, current LLMs still underperform compared to classical classifiers such as SVM and MLP in emotion prediction tasks
We have forgotten the future: cultural memory and the Italian feft's horizon of expectation
No full textThis article analyzes Italy's politics of memory in the age of presentism, a "regime of historicity" that posits a kind of eternal present that corresponds with a neoliberal ethos that naturalizes the current political-economic order. Drawing on mnemonic hegemonic theory, it examines how the loss of a utopian dimension and our problematic relationship with the future influence Italy's flawed relationship with its past. The article focuses on Francesco Piccolo's Il desiderio di essere come tutti (2013), a novel that interweaves autobiography, cultural and historical commentary, and a narrative analysis of the developments of the Italian left from the years (1972-1984) when Enrico Berlinguer was the secretary of the Italian Communist Party (PCI) to the last Berlusconi government (2008-2011). Drawing on the work of Reinhart Koselleck, François Hartog, Enzo Traverso, and others, the article outlines a sociocultural analysis of presentism and of the feeling of "stuckness" that pervades our political conjuncture. Building on this theoretical framework, it then analyzes how Il desiderio di essere come tutti represents the disappearance of the Italian left's horizon of expectation. By working through the historical defeat of the communist left, Piccolo's novel gives us an insight into the causes and effects of presentism in Italian society. This enables me to clarify what I mean by "forgetting the future," and explore how the exhaustion of a particular hope of sociopolitical transformation bears upon Italy's politics of memory. Ultimately, this article aims to open a conversation on whether it is possible to fight multidimensional forgetting in a time in which neoliberal rationality has penetrated every human sphere, remaking the subject and dismantling the social in the process
Not a nuisance but a useful heuristic: outlier dimensions favor frequent tokens in language models
No full textWe study last-layer outlier dimensions, i.e. dimensions that display extreme activations for the majority of inputs. We show that outlier dimensions arise in many different modern language models, and trace their function back to the heuristic of constantly predicting frequent words. We further show how a model can block this heuristic when it is not contextually appropriate, by assigning a counterbalancing weight mass to the remaining dimensions, and we investigate which model parameters boost outlier dimensions and when they arise during training. We conclude that outlier dimensions are a specialized mechanism discovered by many distinct models to implement a useful token prediction heuristic.We thank Beatrix Miranda Ginn Nielsen, Santiago Acevedo, Diego Doimo, Javier Ferrando, Alessandro Laio and the members of the UPF COLT group for feedback. Our work was funded by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 101019291). We also received funding from the Catalan government (AGAUR grant SGR 2021 00470). NG also received the support of a fellowship from Fundación Ramón Areces. GB also received the support of grant PID2020-112602GBI00/MICIN/AEI/10.13039/501100011033, funded by the Ministerio de Ciencia e Innovación and the Agencia Estatal de Investigación (Spain). This paper reflects the authors’ view only, and the funding agencies are not responsible for any use that may be made of the information it contains