Suchergebnisse

International audience ; Numerous domains have interests in studying the viewpoints expressed online, be it for marketing, cybersecurity, or research purposes with the rise of computational social sciences. Current stance detection models are usually grounded on the specificities of some social platforms. This rigidity is unfortunate since it does not allow the integration of the multitude of signals informing effective stance detection. We propose the SCSD model, or Sequential Community-based Stance Detection model, a semi-supervised ensemble algorithm which considers these signals by modeling them as a multi-layer graph representing proximities between profiles. We use a handful of seed profiles, for whom we know the stance, to classify the rest of the profiles by exploiting like-minded communities. These communities represent profiles close enough to assume they share a similar stance on a given subject. Using datasets from two different social platforms, containing two to five stances, we show that by combining several types of proximity we can achieve excellent results. Moreover, we compare the proximities to find those which convey useful information in term of stance detection.

International audience ; Numerous domains have interests in studying the viewpoints expressed online, be it for marketing, cybersecurity, or research purposes with the rise of computational social sciences. Current stance detection models are usually grounded on the specificities of some social platforms. This rigidity is unfortunate since it does not allow the integration of the multitude of signals informing effective stance detection. We propose the SCSD model, or Sequential Community-based Stance Detection model, a semi-supervised ensemble algorithm which considers these signals by modeling them as a multi-layer graph representing proximities between profiles. We use a handful of seed profiles, for whom we know the stance, to classify the rest of the profiles by exploiting like-minded communities. These communities represent profiles close enough to assume they share a similar stance on a given subject. Using datasets from two different social platforms, containing two to five stances, we show that by combining several types of proximity we can achieve excellent results. Moreover, we compare the proximities to find those which convey useful information in term of stance detection.

International audience ; The French presidential election was one of the main political event of 2017, and triggered a lot of activity on Twitter. The campaign was highly unpredictable and led to the rise of 5 main parties instead of the historical bipartite (left-right) confrontation, ranging from far-left to far-right. This dataset paper proposes #Élysée2017fr, a large and complex dataset of 22853 Twitter profiles active during the campaign (from November 2016 to May 2017), and their corresponding tweets and retweets, plus the retweet and mention networks related to these profiles. The profiles were manually annotated with their political affiliations (up to 2 political parties per profile), their nature (individual or collective), and the sex of the profile's owner when available. This is one of the rare datasets that considers a non-binary stance classification and, to our knowledge, the first one with a large number of profiles, and the first one proposing overlapping political communities. This dataset can be used as-is to study the campaign mechanisms on Twitter, or used to test stance detection models or network analysis tools. Mining these data might reveal new insights on current issues like echo chambers or fake news diffusion.

International audience ; The French presidential election was one of the main political event of 2017, and triggered a lot of activity on Twitter. The campaign was highly unpredictable and led to the rise of 5 main parties instead of the historical bipartite (left-right) confrontation, ranging from far-left to far-right. This dataset paper proposes #Élysée2017fr, a large and complex dataset of 22853 Twitter profiles active during the campaign (from November 2016 to May 2017), and their corresponding tweets and retweets, plus the retweet and mention networks related to these profiles. The profiles were manually annotated with their political affiliations (up to 2 political parties per profile), their nature (individual or collective), and the sex of the profile's owner when available. This is one of the rare datasets that considers a non-binary stance classification and, to our knowledge, the first one with a large number of profiles, and the first one proposing overlapping political communities. This dataset can be used as-is to study the campaign mechanisms on Twitter, or used to test stance detection models or network analysis tools. Mining these data might reveal new insights on current issues like echo chambers or fake news diffusion.

International audience ; Numerous domains have interests in studying the viewpoints expressed online, be it for marketing, cybersecurity, or research purposes with the rise of computational social sciences. Current stance detection models are usually grounded on the specificities of some social platforms. This rigidity is unfortunate since it does not allow the integration of the multitude of signals informing effective stance detection. We propose the SCSD model, or Sequential Community-based Stance Detection model, a semi-supervised ensemble algorithm which considers these signals by modeling them as a multi-layer graph representing proximities between profiles. We use a handful of seed profiles, for whom we know the stance, to classify the rest of the profiles by exploiting like-minded communities. These communities represent profiles close enough to assume they share a similar stance on a given subject. Using datasets from two different social platforms, containing two to five stances, we show that by combining several types of proximity we can achieve excellent results. Moreover, we compare the proximities to find those which convey useful information in term of stance detection.

International audience ; Numerous domains have interests in studying the viewpoints expressed online, be it for marketing, cybersecurity, or research purposes with the rise of computational social sciences. Current stance detection models are usually grounded on the specificities of some social platforms. This rigidity is unfortunate since it does not allow the integration of the multitude of signals informing effective stance detection. We propose the SCSD model, or Sequential Community-based Stance Detection model, a semi-supervised ensemble algorithm which considers these signals by modeling them as a multi-layer graph representing proximities between profiles. We use a handful of seed profiles, for whom we know the stance, to classify the rest of the profiles by exploiting like-minded communities. These communities represent profiles close enough to assume they share a similar stance on a given subject. Using datasets from two different social platforms, containing two to five stances, we show that by combining several types of proximity we can achieve excellent results. Moreover, we compare the proximities to find those which convey useful information in term of stance detection.

International audience ; Stance detection systems often integrate social clues in their algorithms. While the influence of social groups on stance is known, there is no evaluation of how well state-of-the-art community detection algorithms perform in terms of detecting like-minded communities, i.e. communities that share the same stance on a given subject. We used Twitter's social interactions to compare the results of community detection algorithms on datasets on the Scottish Independence Referendum and US Midterm Elections. Our results show that algorithms relying on information diffusion perform better for this task and confirm previous observations about retweets being better vectors of stance than mentions.

National audience ; [FR] De nombreux domaines ont intérêt à étudier les points de vue exprimés en ligne, que ce soit le marketing, la cybersécurité, ou les humanités numériques. Nous proposons un modèle générique permettant de détecter ces points de vue en s'appuyant sur la proximité entre les profils utilisateurs, déterminée en fonction des éléments disponibles sur la plateforme étudiée. Nous utilisons un échantillon de profils annotés au départ, et des graphes modélisant la proximité des profils, afin de catégoriser le reste de la population. Ce modèle exploite les communautés, représentant des profils assez proches pour supposer qu'ils partagent une posture similaire. En utilisant plusieurs jeux de données, nous montrons qu'en combinant plusieurs types de proximités, nous obtenons des résultats très pertinents malgré la simplicité de la méthode. [EN] Numerous domains have interests in studying the viewpoints expressed online, be it for marketing, cybersecurity, or computational social sciences. We propose a generic model to detect said viewpoints, relying on the proximity between user profiles, determined according to the information available online. We use a sample of seed profiles for whom we know the stance, and graphs modeling the proximity of the user profiles, to classify the rest of the population. This model exploits communities, which represent profiles close enough to assume they share a similar stance. Using several datasets, we show that by combining several types of proximity we can achieve excellent effectiveness despite the simplicity of the method.

Stance detection systems often integrate social clues in their algorithms. While the influence of social groups on stance is known, there is no evaluation of how well state-of-the-art community detection algorithms perform in terms of detecting like-minded communities, i.e. communities that share the same stance on a given subject. We used Twitter's social interactions to compare the results of community detection algorithms on datasets on the Scottish Independence Referendum and US Midterm Elections. Our results show that algorithms relying on information diffusion perform better for this task and confirm previous observations about retweets being better vectors of stance than mentions.

[FR] De nombreux domaines ont intérêt à étudier les points de vue exprimés en ligne, que ce soit le marketing, la cybersécurité, ou les humanités numériques. Nous proposons un modèle générique permettant de détecter ces points de vue en s'appuyant sur la proximité entre les profils utilisateurs, déterminée en fonction des éléments disponibles sur la plateforme étudiée. Nous utilisons un échantillon de profils annotés au départ, et des graphes modélisant la proximité des profils, afin de catégoriser le reste de la population. Ce modèle exploite les communautés, représentant des profils assez proches pour supposer qu'ils partagent une posture similaire. En utilisant plusieurs jeux de données, nous montrons qu'en combinant plusieurs types de proximités, nous obtenons des résultats très pertinents malgré la simplicité de la méthode. [EN] Numerous domains have interests in studying the viewpoints expressed online, be it for marketing, cybersecurity, or computational social sciences. We propose a generic model to detect said viewpoints, relying on the proximity between user profiles, determined according to the information available online. We use a sample of seed profiles for whom we know the stance, and graphs modeling the proximity of the user profiles, to classify the rest of the population. This model exploits communities, which represent profiles close enough to assume they share a similar stance. Using several datasets, we show that by combining several types of proximity we can achieve excellent effectiveness despite the simplicity of the method.

National audience ; [FR] De nombreux domaines ont intérêt à étudier les points de vue exprimés en ligne, que ce soit le marketing, la cybersécurité, ou les humanités numériques. Nous proposons un modèle générique permettant de détecter ces points de vue en s'appuyant sur la proximité entre les profils utilisateurs, déterminée en fonction des éléments disponibles sur la plateforme étudiée. Nous utilisons un échantillon de profils annotés au départ, et des graphes modélisant la proximité des profils, afin de catégoriser le reste de la population. Ce modèle exploite les communautés, représentant des profils assez proches pour supposer qu'ils partagent une posture similaire. En utilisant plusieurs jeux de données, nous montrons qu'en combinant plusieurs types de proximités, nous obtenons des résultats très pertinents malgré la simplicité de la méthode. [EN] Numerous domains have interests in studying the viewpoints expressed online, be it for marketing, cybersecurity, or computational social sciences. We propose a generic model to detect said viewpoints, relying on the proximity between user profiles, determined according to the information available online. We use a sample of seed profiles for whom we know the stance, and graphs modeling the proximity of the user profiles, to classify the rest of the population. This model exploits communities, which represent profiles close enough to assume they share a similar stance. Using several datasets, we show that by combining several types of proximity we can achieve excellent effectiveness despite the simplicity of the method.

International audience ; Stance detection systems often integrate social clues in their algorithms. While the influence of social groups on stance is known, there is no evaluation of how well state-of-the-art community detection algorithms perform in terms of detecting like-minded communities, i.e. communities that share the same stance on a given subject. We used Twitter's social interactions to compare the results of community detection algorithms on datasets on the Scottish Independence Referendum and US Midterm Elections. Our results show that algorithms relying on information diffusion perform better for this task and confirm previous observations about retweets being better vectors of stance than mentions.

National audience ; [FR] De nombreux domaines ont intérêt à étudier les points de vue exprimés en ligne, que ce soit le marketing, la cybersécurité, ou les humanités numériques. Nous proposons un modèle générique permettant de détecter ces points de vue en s'appuyant sur la proximité entre les profils utilisateurs, déterminée en fonction des éléments disponibles sur la plateforme étudiée. Nous utilisons un échantillon de profils annotés au départ, et des graphes modélisant la proximité des profils, afin de catégoriser le reste de la population. Ce modèle exploite les communautés, représentant des profils assez proches pour supposer qu'ils partagent une posture similaire. En utilisant plusieurs jeux de données, nous montrons qu'en combinant plusieurs types de proximités, nous obtenons des résultats très pertinents malgré la simplicité de la méthode. [EN] Numerous domains have interests in studying the viewpoints expressed online, be it for marketing, cybersecurity, or computational social sciences. We propose a generic model to detect said viewpoints, relying on the proximity between user profiles, determined according to the information available online. We use a sample of seed profiles for whom we know the stance, and graphs modeling the proximity of the user profiles, to classify the rest of the population. This model exploits communities, which represent profiles close enough to assume they share a similar stance. Using several datasets, we show that by combining several types of proximity we can achieve excellent effectiveness despite the simplicity of the method.

International audience ; Stance detection systems often integrate social clues in their algorithms. While the influence of social groups on stance is known, there is no evaluation of how well state-of-the-art community detection algorithms perform in terms of detecting like-minded communities, i.e. communities that share the same stance on a given subject. We used Twitter's social interactions to compare the results of community detection algorithms on datasets on the Scottish Independence Referendum and US Midterm Elections. Our results show that algorithms relying on information diffusion perform better for this task and confirm previous observations about retweets being better vectors of stance than mentions.

National audience ; [FR] De nombreux domaines ont intérêt à étudier les points de vue exprimés en ligne, que ce soit le marketing, la cybersécurité, ou les humanités numériques. Nous proposons un modèle générique permettant de détecter ces points de vue en s'appuyant sur la proximité entre les profils utilisateurs, déterminée en fonction des éléments disponibles sur la plateforme étudiée. Nous utilisons un échantillon de profils annotés au départ, et des graphes modélisant la proximité des profils, afin de catégoriser le reste de la population. Ce modèle exploite les communautés, représentant des profils assez proches pour supposer qu'ils partagent une posture similaire. En utilisant plusieurs jeux de données, nous montrons qu'en combinant plusieurs types de proximités, nous obtenons des résultats très pertinents malgré la simplicité de la méthode. [EN] Numerous domains have interests in studying the viewpoints expressed online, be it for marketing, cybersecurity, or computational social sciences. We propose a generic model to detect said viewpoints, relying on the proximity between user profiles, determined according to the information available online. We use a sample of seed profiles for whom we know the stance, and graphs modeling the proximity of the user profiles, to classify the rest of the population. This model exploits communities, which represent profiles close enough to assume they share a similar stance. Using several datasets, we show that by combining several types of proximity we can achieve excellent effectiveness despite the simplicity of the method.