Hasil untuk "Comparative grammar"

Junyoung Koh, Jaeyun Lee, Soo Yong Kim et al.

Recent work on music question answering (Music-QA) has primarily focused on single-track understanding, where models answer questions about an individual audio clip using its tags, captions, or metadata. However, listeners often describe music in comparative terms, and existing benchmarks do not systematically evaluate reasoning across multiple tracks. Building on the Jamendo-QA dataset, we introduce Jamendo-MT-QA, a dataset and benchmark for multi-track comparative question answering. From Creative Commons-licensed tracks on Jamendo, we construct 36,519 comparative QA items over 12,173 track pairs, with each pair yielding three question types: yes/no, short-answer, and sentence-level questions. We describe an LLM-assisted pipeline for generating and filtering comparative questions, and benchmark representative audio-language models using both automatic metrics and LLM-as-a-Judge evaluation.

Tereza Constantinou, Oliver Shorttle, Miles Cranmer et al.

The discovery of inhabited exoplanets hinges on identifying biosignature gases. JWST can reveal biosignature gases, though current discoveries have yet to evidence life. The central challenge is attribution: how can we confidently identify biogenic sources while ruling out, or deeming unlikely, abiotic explanations? Attribution is particularly difficult for individual planets, especially given the stochastic abiotic processes that can set atmospheric conditions. To address this, we propose a comparative multi-planet approach centred on systemic retrievals: the analysis of multiple planets within a system to empirically define the `abiotic baseline'. This baseline, constructed from obligate uninhabited planets, serves as a local reference point. Systemic retrievals enable marginalisation over inaccessible, latent, shared abiotic parameters within planet evolution models. This is possible because planets within a system are linked by their birth in the same natal disk, have been irradiated by the same evolving star, and have a linked dynamical history. Observations aligning with the abiotic baseline, where the locally-informed abiotic planet evolution models demonstrate high out-of-sample predictive accuracy, are likely non-biological. Potentially biological anomalies are identified as statistical outliers from the abiotic baseline using Bayesian leave-one-out cross-validation. A comparative biosignature is thus defined: an anomaly where a biotic planetary evolution model provides a superior fit than its abiotic counterpart. Where both abiotic and biotic models yield poor predictive accuracy, the anomaly is flagged as an ``unknown unknown"; a signature of either unconstrained abiotic processes, or life as we don't yet know it.

The CHROMA Collaboration, R. Kord Valeshabadi, S. Rezaie et al.

We perform an in-depth comparative analysis of unintegrated parton distribution function (UPDF) models for isolated prompt diphoton production in proton-proton collisions at $\sqrt{s}=13$~TeV within the $k_t$-factorization framework. Predictions are obtained with three UPDF approaches: Parton Branching (PB), NLO-MRW, and Modified KMRW (MKMRW). Tree-level $q + \bar q\!\to\!γ+γ$, $q + \bar q\!\to\!γ+ γ+ g$, and $q + g\!\to\!γ+γ+ q$ subprocesses are generated with \textsc{KaTie} using off-shell initial states; the loop-induced $g + g\!\to\!γ+ γ$ channel is evaluated independently. We compare differential cross sections with ATLAS measurements across a broad set of observables, including the photon transverse momenta, diphoton invariant mass and transverse momentum, the Collins-Soper angle, acoplanarity, $φ^*_η$, and a transverse thrust-related variable. This comparative study quantifies the impact of the UPDF choice on the diphoton spectra. We find that the PB model provides the most consistent agreement over all distributions, whereas NLO-MRW tends to overshoot in regions correlated with larger factorization scales and MKMRW generally undershoots due to stronger Sudakov suppression. With standard scale variations, our results indicate that $k_t$-factorization with PB UPDFs can accurately describe diphoton production, while fixed-order collinear predictions typically require higher-order corrections together with parton-shower effects to achieve a comparable description.

Zhanyi Jiao, Qiuqi Wang, Yimiao Zhao

Backtesting risk measures is a central task in financial regulation. While standard backtests evaluate whether a forecasting model is statistically consistent with observed losses, regulatory practice often requires assessing the performance of an internal model relative to benchmark models. We develop a non-parametric sequential framework for comparative backtests of general elicitable risk measures using e-values and e-processes. The proposed methods provide anytime-valid inference and remain robust under dependence and model misspecification. In particular, we propose a modified three-zone approach based on weak dominance, which yields more informative conclusions in comparative backtesting. As a technical building block, we also construct general standard e-backtests for identifiable risk measures and characterize the associated e-values and e-processes. The resulting procedures apply to a broad class of commonly used risk measures, including the mean, variance, Value-at-Risk, Expected Shortfall, and expectiles. Simulation studies and empirical analyses illustrate the effectiveness of the proposed approach.

Gora Zaragoza Ninet

Sugerimos un marco teórico-práctico para el análisis de textos (traducidos) desde una perspectiva feminista; un modelo integrador que el alumnado pueda aplicar en asignaturas afines de grado y postgrado, colaborando de esta forma a la incorporación de la perspectiva de género en la docencia de la traducción. Sugeriremos una base teórica que parte de los estudios de género y traducción para destacar conceptos clave para un análisis traductológico con perspectiva feminista, tales como el estudio de la re-traducción, la recepción, las estrategias de traducción (feministas) textuales y paratextuales, así como de los agentes implicados en el proceso de traducción.

Quratulain ., Tanzeel Ur Rehman, Bisma Butt

The study focuses on the development of semantic processing in children through the lens of the Disney animated movie “inside out” by Pete Docter and Ronnie Del Carmen. Semantic processing is a basic aspect of human knowledge of language processing that enables to extract meaning from the words and to comprehend the intended information conveyed by others. The movie “Inside Out” depicts the story of an 11-year-old girl named Riley and the personification of her emotion Joy, Sadness, Anger, Fear and Disgust as they navigate her life experiences. By analyzing the movie’s plot, characters and dialogues, the study investigates (1) how children’s semantic processing develops over time, (2) how it relates to the ability to understand abstract concepts and how it can impact children’s emotional development. (3) What role do metaphors and analogies play in the semantic development of children depicted in the movie. The goals of this study are to investigate the content, characters, and themes of the movie, to analyze the use of metaphors and analogies in the movie and to explore the relationship between the development of semantic processing and children's emotional advancement. The present study used mixed method to provide a comprehensive understanding of the phenomenon. The study targeted two different age groups i.e. 5-8 years and 9-12 years to calculate the impact of age on the development of semantic processing. The findings indicate that the movie “Inside Out” successfully engaged the children in reflecting on and succussing their understanding of emotions. Participants demonstrated and enhanced awareness of the importance and complexity of emotions. Also, it provides insight into how animated movies can be used as a tool to promote semantic development in children’s emotional advancement and inform the development of educational materials that can facilitate semantic processing in children.

Fatima Ambreen, Burhan Ozdemir, Dina Abdel Salam El-Dakhs

Abstract Textism, which refers to the special language used in text messaging, is widely used in several forms of communication, such as texting, online chatting and emails. This increased use of textism has alarmed the media and the public because this special language can impact the academic writing of the younger generation. This paper aimed at exploring the attitudes of undergraduate Saudi learners of English as a foreign language (EFL) toward using textism in their academic writing. (115) students of different levels of language proficiency completed a survey about the use of textism across several modalities and recipients. Additionally, (240) student essays were analyzed to identify the potential use of textism in student writing. The results showed that students are generally conscious of the appropriate contexts for the use of textism. However, when it comes to the actual use of textism in writing, low-proficient students tend to produce more textism than those who are at medium and high levels of writing competence.

Owen Henkel, Libby Hills

Machine Learning models have many potentially beneficial applications in education settings, but a key barrier to their development is securing enough data to train these models. Labelling educational data has traditionally relied on highly skilled raters using complex, multi-class rubrics, making the process expensive and difficult to scale. An alternative, more scalable approach could be to use non-expert crowdworkers to evaluate student work, however, maintaining sufficiently high levels of accuracy and inter-rater reliability when using non-expert workers is challenging. This paper reports on two experiments investigating using non-expert crowdworkers and comparative judgement to evaluate complex student data. Crowdworkers were hired to evaluate student responses to open-ended reading comprehension questions. Crowdworkers were randomly assigned to one of two conditions: the control, where they were asked to decide whether answers were correct or incorrect (i.e., a categorical judgement), or the treatment, where they were shown the same question and answers, but were instead asked to decide which of two candidate answers was more correct (i.e., a comparative/preference-based judgement). We found that using comparative judgement substantially improved inter-rater reliability on both tasks. These results are in-line with well-established literature on the benefits of comparative judgement in the field of educational assessment, as well as with recent trends in artificial intelligence research, where comparative judgement is becoming the preferred method for providing human feedback on model outputs when working with non-expert crowdworkers. However, to our knowledge, these results are novel and important in demonstrating the beneficial effects of using the combination of comparative judgement and crowdworkers to evaluate educational data.

Danna Pinto, Anat Prior, Elana Zion Golumbic

Ya-mei Chen

The Translation-Editing-Proofreading (TEP) model is an established approach for ensuring the quality of translation services in the translation industry. However, when the TEP process is replicated in a crowdsourcing context, new factors influence its effectiveness. To address this issue, this article, based on Engeström's activity theory, provides a theoretical account of those factors that influence the TEP process reproduced in Global Voices Lingua. The article then presents two interrelated empirical studies on (1) Lingua volunteers' profiles and motivations and (2) Chinese Lingua editors' revisions to explain how to examine whether unfavourable influences exist in practice. The results reveal that for the Lingua project as a whole, the volunteers' ability and motivational tendencies do not produce substantial adverse effects. Likewise, the Chinese revisions seem to be adequate to guarantee the desired quality although Lingua's language communities may display different scenarios. Further empirical investigation of other aspects (e.g. the adequacy of the translator's work and translator-editor interactions) is still needed to evaluate the replication of the TEP model in Global Voices Lingua more precisely.

R. G. Seymour, A. Nyarko-Agyei, H. R. McCabe et al.

Forcing someone into marriage against their will is a violation of their human rights. In 2021, the county of Nottinghamshire, UK, launched a strategy to tackle forced marriage and violence against women and girls. However, accessing information about where victims are located in the county could compromise their safety, so it is not possible to develop interventions for different areas of the county. Comparative judgement studies offer a way to map the risk of human rights abuses without collecting data that could compromise victim safety. Current methods require studies to have a large number of participants, so we develop a comparative judgement model that provides a more flexible spatial modelling structure and a mechanism to schedule comparisons more effectively. The methods reduce the data collection burden on participants and make a comparative judgement study feasible with a small number of participants. Underpinning these methods is a latent variable representation that improves on the scalability of previous comparative judgement models. We use these methods to map the risk of forced marriage across Nottinghamshire thereby supporting the county's strategy for tackling violence against women and girls.

Ares Fisher, Rajesh P. N. Rao

Human vision involves parsing and representing objects and scenes using structured representations based on part-whole hierarchies. Computer vision and machine learning researchers have recently sought to emulate this capability using capsule networks, reference frames and active predictive coding, but a generative model formulation has been lacking. We introduce Recursive Neural Programs (RNPs), which, to our knowledge, is the first neural generative model to address the part-whole hierarchy learning problem. RNPs model images as hierarchical trees of probabilistic sensory-motor programs that recursively reuse learned sensory-motor primitives to model an image within different reference frames, forming recursive image grammars. We express RNPs as structured variational autoencoders (sVAEs) for inference and sampling, and demonstrate parts-based parsing, sampling and one-shot transfer learning for MNIST, Omniglot and Fashion-MNIST datasets, demonstrating the model's expressive power. Our results show that RNPs provide an intuitive and explainable way of composing objects and scenes, allowing rich compositionality and intuitive interpretations of objects in terms of part-whole hierarchies.

Lunjia Hu, Charlotte Peale

In many learning theory problems, a central role is played by a hypothesis class: we might assume that the data is labeled according to a hypothesis in the class (usually referred to as the realizable setting), or we might evaluate the learned model by comparing it with the best hypothesis in the class (the agnostic setting). Taking a step beyond these classic setups that involve only a single hypothesis class, we introduce comparative learning as a combination of the realizable and agnostic settings in PAC learning: given two binary hypothesis classes $S$ and $B$, we assume that the data is labeled according to a hypothesis in the source class $S$ and require the learned model to achieve an accuracy comparable to the best hypothesis in the benchmark class $B$. Even when both $S$ and $B$ have infinite VC dimensions, comparative learning can still have a small sample complexity. We show that the sample complexity of comparative learning is characterized by the mutual VC dimension $\mathsf{VC}(S,B)$ which we define to be the maximum size of a subset shattered by both $S$ and $B$. We also show a similar result in the online setting, where we give a regret characterization in terms of the mutual Littlestone dimension $\mathsf{Ldim}(S,B)$. These results also hold for partial hypotheses. We additionally show that the insights necessary to characterize the sample complexity of comparative learning can be applied to characterize the sample complexity of realizable multiaccuracy and multicalibration using the mutual fat-shattering dimension, an analogue of the mutual VC dimension for real-valued hypotheses. This not only solves an open problem proposed by Hu, Peale, Reingold (2022), but also leads to independently interesting results extending classic ones about regression, boosting, and covering number to our two-hypothesis-class setting.

Renato Cândido da Silva, Orlando Luiz de Araújo

No Brasil, o mito de Electra foi de interesse de diversos dramaturgos, como no caso de Ivo Bender, ao escrever Bodas ao cair da tarde (1983). Nessa peça, a fim de atender aos moldes de um texto breve de teatro, o mito grego é “reduzido”, condensado em poucas páginas. Mas também é “ampliado”, possibilitando uma nova leitura do conhecido mito da tradição clássica. Com esse propósito, o artigo visa a analisar o mito de Electra presente em Bodas ao cair da tarde, levando em considerações as principais mudanças e inovações empreendidas por Ivo Bender no processo de reescritura.

Raghavendra Sridharamurthy, Vijay Natarajan

Comparative analysis of scalar fields is an important problem with various applications including feature-directed visualization and feature tracking in time-varying data. Comparing topological structures that are abstract and succinct representations of the scalar fields lead to faster and meaningful comparison. While there are many distance or similarity measures to compare topological structures in a global context, there are no known measures for comparing topological structures locally. While the global measures have many applications, they do not directly lend themselves to fine-grained analysis across multiple scales. We define a local variant of the tree edit distance and apply it towards local comparative analysis of merge trees with support for finer analysis. We also present experimental results on time-varying scalar fields, 3D cryo-electron microscopy data, and other synthetic data sets to show the utility of this approach in applications like symmetry detection and feature tracking.

Giovanna Marina Alfonzetti

En el presente artículo se analiza un corpus de pintadas con insultos fotografiadas en los recintos del Departamento de Ciencias Humanísticas (Catania) y la Sapienza (Roma) en dos fases: 2007-2010 y 2019. Con ello, se demuestra que la expansión de las redes sociales ha reducido drásticamente la producción de pintadas de todo tipo. Los insultos, indicios significativos para reconstruir el sistema de valores de una comunidad sociocultural, se entienden como actos que atacan directamente al receptor. Así pues, se sitúan en el marco de la violencia verbal, coincidiendo parcialmente con otros actos violentos, de los que tratamos de diferenciarlos en base a una serie de criterios (pragmático-funcionales, formales y secuenciales), tal y como prevé la teoría del prototipo. En relación con la (des)cortesía verbal, los insultos prototípicos podrían considerarse actos intrínsecamente maleducados, cuya expresión presenta una clara tendencia a agravarse en varios niveles. Por consiguiente, se propone una clasificación de los insultos convencionales del corpus, siendo en muchos casos el resultado de procesos de metaforización.

Mariana Braga

Este artigo pretende analisar possíveis pontos de interseção semânticos e estruturais entre o género epistolar e o género ensaístico na obra crítica de Helder Macedo. Tomarei como objeto de estudo Camões e outros contemporâneos (2017), sua mais nova coletânea de textos críticos revisitados, a qual inclui quatro ensaios sobre Luís de Camões. Entre eles destaca-se “Camões: o imaginário da malandragem”, em que o ensaísta se debruça sobre cartas cuja autoria é atribuída ao poeta. Para refletir sobre a importância do amor na obra de Helder Macedo – como tema, força motriz e modus operandi –, será fundamental o estudo de semiótica Presenças do outro (2002), de Eric Landowski, no qual há um capítulo dedicado especificamente às cartas de amor, entendidas como um meio pelo qual se procura tornar presente o outro ausente, ou seja, como ato de presentificação do outro.

Mona Valipoor

The Arabic grammar tradition is one of the most important linguistic schools in history- a school whose achievements in many cases are as important as the achievements of the modern linguistic studies. The present study, by explaining the conformity of the threefold divisions of the verbal mood (declarative, subjunctive, and imperative) with the triple subdivision of the verbs into marfu’, mansub and majzum in Arabic grammar, attempts to explain how the Arabic language grammarians have described the functions of subjunctive mood through the elements they have referred to. The results of this study indicate that Sibawayhi, the leader of the school of Basra and the writer of the oldest book on the Arabic grammar, has been aware of the role of realis/irrealis distinction in the uses of the mansub (subjunctive) and marfu’ (declarative). Moreover, through some of his comments, it is possible to derive the importance of the speaker’s view in determining whether or not the proposition is definite. On the other hand, the study reveals how al-Farra, the most prominent grammarian of the Kufa School, has emphasized the role of time and temporal implications in explaining the applications of the subjunctive mood. Subsequent grammarians, following their syntactic views, have supplemented this discussion and, for instance, have mentioned the distinction between absolute time and relative time to determine temporal implications of the present subjunctive mood. Currently, linguists without having access to the findings of this tradition, have discovered the importance of these two components in the applications of the subjunctive mood; On the basis of the findings of the present study, such an achievement can be traced its origin back to more than a thousand years ago in the books of Arabic grammarians.

Irfan Shah

Shanul Haq Haqqee (15th September, 1917 - 11th October, 2005) was a versatile genius having multi-dimensional vision and knowledge. He was at a time great linguist, lexicographer, researcher, scholar, critic, translator, biographer, fiction writer, an acknowledged great poet, a story writer for children, humorist, copywriter and a publicist. His respected father Moulvi Ahteshamuddin was also a great lexicographer, poet, researcher and writer of great eminence. Shanul Haq Haqqee's creative works is found in various forms of Urdu literature in which he has shown his creativity and exhibited his new experiences in them. In his works his approach and vision had various angles and innovations. His writings in respect of credibility and correctness of language attained status of a certificate of perfection. His special interest in lexicography represents his command on the subject and theme. During his honorary appointment in Urdu Development Board, Karachi, he served for many noble services which includes his lexicographic work on Urdu dictionary and issuance of quarterly "Urdu Nama". He also remained associated as Chief Editor Urdu Monthly "Mah-e-Nau" Karachi.

Talía Vázquez Romaguera, Liset Vázquez Romaguera, David Castro Piñol et al.

In the last decade, the development of technologies and tools for eye tracking has been a constantly growing area. Detecting the center of the pupil, using image processing techniques, has been an essential step in this process. A large number of techniques have been proposed for pupil center detection using both traditional image processing and machine learning-based methods. Despite the large number of methods proposed, no comparative work on their performance was found, using the same images and performance metrics. In this work, we aim at comparing four of the most frequently cited traditional methods for pupil center detection in terms of accuracy, robustness, and computational cost. These methods are based on the circular Hough transform, ellipse fitting, Daugman's integro-differential operator and radial symmetry transform. The comparative analysis was performed with 800 infrared images from the CASIA-IrisV3 and CASIA-IrisV4 databases containing various types of disturbances. The best performance was obtained by the method based on the radial symmetry transform with an accuracy and average robustness higher than 94%. The shortest processing time, obtained with the ellipse fitting method, was 0.06 s.