1,720,956 research outputs found
Traitement automatique de la langue dans le contexte de la réponse aux appels d’offres complexes
No full textThe tender analysis process, characterized by complex document evaluationsunder tight deadlines, is often resource-intensive and prone to errors. Keyphases of this process rely on Hierarchical Multi-label Text Classification(HMTC) tasks. The goal is to classify a given input document from the tendercorpora into labels derived from a taxonomic hierarchy.While discriminative methods have traditionally dominated HMTC literature,generative models have shown significant promise as alternatives. This thesisdelves deeper into generative methods to analyze their ability to learn both,the hierarchical information and the semantic relationships among labels.Notably, most generative methods perform well at preserving the class hier-archy when making predictions which diminishes concerns over hierarchicalintegrity. However, by design, these solutions set aside what made thestrength of discriminative methods, that is to make predictions based on theentire hierarchical information and to capture correlation from input textand labels for better classifications.We then, introduce a hybrid approach that leverages the strengths of bothmethodologies through a non-autoregressive decoder enhanced with a novelHierarchical Self-Attention mechanism. The decoder input is expanded withlabel embedding. Harnessing the power of the Cross-Attention and Hier-archical Self-Attention mechanisms, we achieve a label representation thatbenefits from instance and global label-wise information. This mechanismfacilitates simultaneous decoding of the entire hierarchy, utilizing global andbidirectional attention alongside labels. Through a set of experimental stud-ies on several large scale datasets, we demonstrate that this design deliverssuperior performance compared to 16 state-of-the-art HMTC models.L'analyse des appels d'offres, qui implique l'évaluation de documents complexes, constitue un processus exigeant en ressources et susceptible de générer des erreurs en raison de sa nature intrinsèquement laborieuse et des contraintes temporelles imposées. Deux phases fondamentales de ce processus reposent sur des tâches de classification hiérarchique multi-label du texte (CHMT), dont l'objectif est d'attribuer à chaque document du corpus d'appels d'offres des catégories spécifiques, organisées au sein d'une taxonomie hiérarchique prédéfinie.Historiquement, les méthodes discriminatives ont dominé la littérature. Toutefois, les modèles génératifs ont récemment émergé comme des alternatives prometteuses. Cette thèse explore en profondeur les méthodes génératives afin d'analyser leur capacité à apprendre à la fois les informations hiérarchiques et les relations sémantiques entre les classes.En particulier, la plupart des méthodes génératives se distinguent par leur aptitude à préserver l'intégrité de la hiérarchie lors des prédictions. Cependant, en raison de leur conception, ces approches ne parviennent pas à exploiter pleinement les informations hiérarchiques ni à capturer efficacement les corrélations fines entre le texte d'entrée et les catégories, limitant ainsi leur capacité à atteindre de meilleures performances en classification.Nous proposons ainsi une approche hybride qui combine les atouts des deux méthodologies, en intégrant un décodeur non auto-régressif, renforcé par un nouveau mécanisme d'attention hiérarchique. L'entrée du décodeur est pré-peuplée par des représentations vectorielles de tous les nœuds de la taxonomie, fournissant ainsi une base complète pour guider le processus de décodage. En exploitant les mécanismes d'attention croisée et d'attention hiérarchique, nous obtenons une représentation des classes qui fusionne à la fois des informations spécifiques aux instances de texte et des informations provenant de la structure globale de la taxonomie. Ces mécanismes permettent donc un décodage simultané de l'ensemble de la hiérarchie.À travers une série d'expérimentations sur plusieurs jeux de données, nous montrons que cette approche surpasse 16 modèles de l'état de l'art en CHMT
Traitement automatique de la langue dans le contexte de la réponse aux appels d’offres complexes
No full textThe tender analysis process, characterized by complex document evaluationsunder tight deadlines, is often resource-intensive and prone to errors. Keyphases of this process rely on Hierarchical Multi-label Text Classification(HMTC) tasks. The goal is to classify a given input document from the tendercorpora into labels derived from a taxonomic hierarchy.While discriminative methods have traditionally dominated HMTC literature,generative models have shown significant promise as alternatives. This thesisdelves deeper into generative methods to analyze their ability to learn both,the hierarchical information and the semantic relationships among labels.Notably, most generative methods perform well at preserving the class hier-archy when making predictions which diminishes concerns over hierarchicalintegrity. However, by design, these solutions set aside what made thestrength of discriminative methods, that is to make predictions based on theentire hierarchical information and to capture correlation from input textand labels for better classifications.We then, introduce a hybrid approach that leverages the strengths of bothmethodologies through a non-autoregressive decoder enhanced with a novelHierarchical Self-Attention mechanism. The decoder input is expanded withlabel embedding. Harnessing the power of the Cross-Attention and Hier-archical Self-Attention mechanisms, we achieve a label representation thatbenefits from instance and global label-wise information. This mechanismfacilitates simultaneous decoding of the entire hierarchy, utilizing global andbidirectional attention alongside labels. Through a set of experimental stud-ies on several large scale datasets, we demonstrate that this design deliverssuperior performance compared to 16 state-of-the-art HMTC models.L'analyse des appels d'offres, qui implique l'évaluation de documents complexes, constitue un processus exigeant en ressources et susceptible de générer des erreurs en raison de sa nature intrinsèquement laborieuse et des contraintes temporelles imposées. Deux phases fondamentales de ce processus reposent sur des tâches de classification hiérarchique multi-label du texte (CHMT), dont l'objectif est d'attribuer à chaque document du corpus d'appels d'offres des catégories spécifiques, organisées au sein d'une taxonomie hiérarchique prédéfinie.Historiquement, les méthodes discriminatives ont dominé la littérature. Toutefois, les modèles génératifs ont récemment émergé comme des alternatives prometteuses. Cette thèse explore en profondeur les méthodes génératives afin d'analyser leur capacité à apprendre à la fois les informations hiérarchiques et les relations sémantiques entre les classes.En particulier, la plupart des méthodes génératives se distinguent par leur aptitude à préserver l'intégrité de la hiérarchie lors des prédictions. Cependant, en raison de leur conception, ces approches ne parviennent pas à exploiter pleinement les informations hiérarchiques ni à capturer efficacement les corrélations fines entre le texte d'entrée et les catégories, limitant ainsi leur capacité à atteindre de meilleures performances en classification.Nous proposons ainsi une approche hybride qui combine les atouts des deux méthodologies, en intégrant un décodeur non auto-régressif, renforcé par un nouveau mécanisme d'attention hiérarchique. L'entrée du décodeur est pré-peuplée par des représentations vectorielles de tous les nœuds de la taxonomie, fournissant ainsi une base complète pour guider le processus de décodage. En exploitant les mécanismes d'attention croisée et d'attention hiérarchique, nous obtenons une représentation des classes qui fusionne à la fois des informations spécifiques aux instances de texte et des informations provenant de la structure globale de la taxonomie. Ces mécanismes permettent donc un décodage simultané de l'ensemble de la hiérarchie.À travers une série d'expérimentations sur plusieurs jeux de données, nous montrons que cette approche surpasse 16 modèles de l'état de l'art en CHMT
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Dispelling the Myths Behind First-author Citation Counts
Full text linkWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
koamabayili/VECTRON-author-checklist: VECTRON author checklist
No full textWe have done our best to complete the author checklist relating to the use of animals in the hut study. Note that the objective for the hut study was to evaluate the IRS treatment applications for residual efficacy against Anopheles mosquitoes, including the local An. coluzzii mosquito population. Cows were only used to attract mosquitoes into the huts and no tests were carried out directly on the cows. The author checklist is intended for use with studies where experiments are carried out on animals, which is why we have had such difficulty in completing this for the hut study, as many of the questions do not relate to how the cows were used
Author-wise bibliometric analysis based on entropy.
No full textAuthor-wise bibliometric analysis based on entropy.</p
- …
