Cambios
En el instante 21 de octubre de 2025, 8:59:11 UTC,
-
Añadido recurso Decoding Online Hate in the United States: A BERT-CNN Analysis of 36 Million Tweets from 2020 to 2022 a Decoding Online Hate in the United States: A BERT-CNN Analysis of 36 Million Tweets from 2020 to 2022
| f | 1 | { | f | 1 | { |
| 2 | "author": "SS Pandey, A Garcia-Robledo, M Zangiabady", | 2 | "author": "SS Pandey, A Garcia-Robledo, M Zangiabady", | ||
| 3 | "author_email": null, | 3 | "author_email": null, | ||
| 4 | "creator_user_id": "a3da3ec9-3fd4-47a4-8d04-0a90b09614e0", | 4 | "creator_user_id": "a3da3ec9-3fd4-47a4-8d04-0a90b09614e0", | ||
| 5 | "extras": [ | 5 | "extras": [ | ||
| 6 | { | 6 | { | ||
| 7 | "key": "A\u00f1o", | 7 | "key": "A\u00f1o", | ||
| 8 | "value": "2024" | 8 | "value": "2024" | ||
| 9 | }, | 9 | }, | ||
| 10 | { | 10 | { | ||
| 11 | "key": "DOI", | 11 | "key": "DOI", | ||
| 12 | "value": "https://doi.org/10.1109/icsc59802.2024.00059" | 12 | "value": "https://doi.org/10.1109/icsc59802.2024.00059" | ||
| 13 | }, | 13 | }, | ||
| 14 | { | 14 | { | ||
| 15 | "key": "Google Scholar URL", | 15 | "key": "Google Scholar URL", | ||
| 16 | "value": | 16 | "value": | ||
| 17 | gesize=100&sortby=pubdate&citation_for_view=b81TvMMAAAAJ:Se3iqnhoufwC" | 17 | gesize=100&sortby=pubdate&citation_for_view=b81TvMMAAAAJ:Se3iqnhoufwC" | ||
| 18 | }, | 18 | }, | ||
| 19 | { | 19 | { | ||
| 20 | "key": "Identificador hash", | 20 | "key": "Identificador hash", | ||
| 21 | "value": "5f6f16bf7304" | 21 | "value": "5f6f16bf7304" | ||
| 22 | }, | 22 | }, | ||
| 23 | { | 23 | { | ||
| 24 | "key": "Lugar de publicaci\u00f3n", | 24 | "key": "Lugar de publicaci\u00f3n", | ||
| 25 | "value": "2024 IEEE 18th International Conference on Semantic | 25 | "value": "2024 IEEE 18th International Conference on Semantic | ||
| 26 | Computing (ICSC), 329-334, 2024" | 26 | Computing (ICSC), 329-334, 2024" | ||
| 27 | }, | 27 | }, | ||
| 28 | { | 28 | { | ||
| 29 | "key": "Tipo", | 29 | "key": "Tipo", | ||
| 30 | "value": "Publicaci\u00f3n" | 30 | "value": "Publicaci\u00f3n" | ||
| 31 | }, | 31 | }, | ||
| 32 | { | 32 | { | ||
| 33 | "key": "Tipo de publicaci\u00f3n", | 33 | "key": "Tipo de publicaci\u00f3n", | ||
| 34 | "value": "Conferencia" | 34 | "value": "Conferencia" | ||
| 35 | }, | 35 | }, | ||
| 36 | { | 36 | { | ||
| 37 | "key": "URL directo", | 37 | "key": "URL directo", | ||
| 38 | "value": | 38 | "value": | ||
| 39 | "https://research.utwente.nl/files/351518475/Paper_2024.pdf" | 39 | "https://research.utwente.nl/files/351518475/Paper_2024.pdf" | ||
| 40 | } | 40 | } | ||
| 41 | ], | 41 | ], | ||
| 42 | "groups": [ | 42 | "groups": [ | ||
| 43 | { | 43 | { | ||
| 44 | "description": "", | 44 | "description": "", | ||
| 45 | "display_name": "Publicaciones", | 45 | "display_name": "Publicaciones", | ||
| 46 | "id": "8be672a5-4640-455e-a4f3-46b52b66c07b", | 46 | "id": "8be672a5-4640-455e-a4f3-46b52b66c07b", | ||
| 47 | "image_display_url": "", | 47 | "image_display_url": "", | ||
| 48 | "name": "publicaciones", | 48 | "name": "publicaciones", | ||
| 49 | "title": "Publicaciones" | 49 | "title": "Publicaciones" | ||
| 50 | } | 50 | } | ||
| 51 | ], | 51 | ], | ||
| 52 | "id": "2ca1c136-7d06-46e8-8ab7-55b965a55832", | 52 | "id": "2ca1c136-7d06-46e8-8ab7-55b965a55832", | ||
| 53 | "isopen": false, | 53 | "isopen": false, | ||
| 54 | "license_id": null, | 54 | "license_id": null, | ||
| 55 | "license_title": null, | 55 | "license_title": null, | ||
| 56 | "maintainer": null, | 56 | "maintainer": null, | ||
| 57 | "maintainer_email": null, | 57 | "maintainer_email": null, | ||
| 58 | "metadata_created": "2025-10-21T08:59:10.753634", | 58 | "metadata_created": "2025-10-21T08:59:10.753634", | ||
| n | 59 | "metadata_modified": "2025-10-21T08:59:10.753641", | n | 59 | "metadata_modified": "2025-10-21T08:59:11.211465", |
| 60 | "name": "5f6f16bf7304", | 60 | "name": "5f6f16bf7304", | ||
| 61 | "notes": "Since its inception, social media has enabled people | 61 | "notes": "Since its inception, social media has enabled people | ||
| 62 | worldwide to connect with like-minded individuals and freely express | 62 | worldwide to connect with like-minded individuals and freely express | ||
| 63 | their thoughts and opinions. However, its widespread nature has not | 63 | their thoughts and opinions. However, its widespread nature has not | ||
| 64 | only had an immeasurable impact on society but also presented | 64 | only had an immeasurable impact on society but also presented | ||
| 65 | significant challenges. One such challenge is online hate speech. | 65 | significant challenges. One such challenge is online hate speech. | ||
| 66 | Consequently, the identification of hate speech has recently gained | 66 | Consequently, the identification of hate speech has recently gained | ||
| 67 | considerable attention, ranging from reactive methods, such as | 67 | considerable attention, ranging from reactive methods, such as | ||
| 68 | classifying individual posts, to proactive strategies that utilize | 68 | classifying individual posts, to proactive strategies that utilize | ||
| 69 | contextual information to decipher the complex lexicon of online | 69 | contextual information to decipher the complex lexicon of online | ||
| 70 | discussions. Despite these efforts, current research lacks a | 70 | discussions. Despite these efforts, current research lacks a | ||
| 71 | comprehensive analysis of hate speech on Twitter during the crucial | 71 | comprehensive analysis of hate speech on Twitter during the crucial | ||
| 72 | 2020-2022 period, marked by significant events such as the COVID-19 | 72 | 2020-2022 period, marked by significant events such as the COVID-19 | ||
| 73 | pandemic. In this paper, we present a BERT-based model for classifying | 73 | pandemic. In this paper, we present a BERT-based model for classifying | ||
| 74 | hate speech. To this end, we collected 36 million tweets posted in the | 74 | hate speech. To this end, we collected 36 million tweets posted in the | ||
| 75 | United States on Twitter during this period. We developed, trained, | 75 | United States on Twitter during this period. We developed, trained, | ||
| 76 | and tested a BERT-based Convolutional Neural Network (BERT-CNN), using | 76 | and tested a BERT-based Convolutional Neural Network (BERT-CNN), using | ||
| 77 | it to classify the collected tweets. The classification of this | 77 | it to classify the collected tweets. The classification of this | ||
| 78 | dataset revealed a high incidence of targets motivated by ethnicity, | 78 | dataset revealed a high incidence of targets motivated by ethnicity, | ||
| 79 | with gender and nationality as other prominent categories. This work | 79 | with gender and nationality as other prominent categories. This work | ||
| 80 | provides insightful data on the sentiments of individuals across the | 80 | provides insightful data on the sentiments of individuals across the | ||
| 81 | United States during the events of 2020-2022.", | 81 | United States during the events of 2020-2022.", | ||
| n | 82 | "num_resources": 0, | n | 82 | "num_resources": 1, |
| 83 | "num_tags": 4, | 83 | "num_tags": 4, | ||
| 84 | "organization": { | 84 | "organization": { | ||
| 85 | "approval_status": "approved", | 85 | "approval_status": "approved", | ||
| 86 | "created": "2022-05-19T00:10:30.480393", | 86 | "created": "2022-05-19T00:10:30.480393", | ||
| 87 | "description": "Observatorio Metropolitano CentroGeo", | 87 | "description": "Observatorio Metropolitano CentroGeo", | ||
| 88 | "id": "b3b3a79d-748a-4464-9471-732b6c74ec53", | 88 | "id": "b3b3a79d-748a-4464-9471-732b6c74ec53", | ||
| 89 | "image_url": | 89 | "image_url": | ||
| 90 | "2022-05-19-001030.456616FullColor1280x1024LogoOnly.png", | 90 | "2022-05-19-001030.456616FullColor1280x1024LogoOnly.png", | ||
| 91 | "is_organization": true, | 91 | "is_organization": true, | ||
| 92 | "name": "observatorio-metropolitano-centrogeo", | 92 | "name": "observatorio-metropolitano-centrogeo", | ||
| 93 | "state": "active", | 93 | "state": "active", | ||
| 94 | "title": "Observatorio Metropolitano CentroGeo", | 94 | "title": "Observatorio Metropolitano CentroGeo", | ||
| 95 | "type": "organization" | 95 | "type": "organization" | ||
| 96 | }, | 96 | }, | ||
| 97 | "owner_org": "b3b3a79d-748a-4464-9471-732b6c74ec53", | 97 | "owner_org": "b3b3a79d-748a-4464-9471-732b6c74ec53", | ||
| 98 | "private": false, | 98 | "private": false, | ||
| 99 | "relationships_as_object": [], | 99 | "relationships_as_object": [], | ||
| 100 | "relationships_as_subject": [], | 100 | "relationships_as_subject": [], | ||
| t | 101 | "resources": [], | t | 101 | "resources": [ |
| 102 | { | ||||
| 103 | "cache_last_updated": null, | ||||
| 104 | "cache_url": null, | ||||
| 105 | "created": "2025-10-21T08:59:11.230486", | ||||
| 106 | "datastore_active": false, | ||||
| 107 | "description": "Since its inception, social media has enabled | ||||
| 108 | people worldwide to connect with like-minded individuals and freely | ||||
| 109 | express their thoughts and opinions. However, its widespread nature | ||||
| 110 | has not only had an immeasurable impact on society but also presented | ||||
| 111 | significant challenges. One such challenge is online hate speech. | ||||
| 112 | Consequently, the identification of hate speech has recently gained | ||||
| 113 | considerable attention, ranging from reactive methods, such as | ||||
| 114 | classifying individual posts, to proactive strategies that utilize | ||||
| 115 | contextual information to decipher the complex lexicon of online | ||||
| 116 | discussions. Despite these efforts, current research lacks a | ||||
| 117 | comprehensive analysis of hate speech on Twitter during the crucial | ||||
| 118 | 2020-2022 period, marked by significant events such as the COVID-19 | ||||
| 119 | pandemic. In this paper, we present a BERT-based model for classifying | ||||
| 120 | hate speech. To this end, we collected 36 million tweets posted in the | ||||
| 121 | United States on Twitter during this period. We developed, trained, | ||||
| 122 | and tested a BERT-based Convolutional Neural Network (BERT-CNN), using | ||||
| 123 | it to classify the collected tweets. The classification of this | ||||
| 124 | dataset revealed a high incidence of targets motivated by ethnicity, | ||||
| 125 | with gender and nationality as other prominent categories. This work | ||||
| 126 | provides insightful data on the sentiments of individuals across the | ||||
| 127 | United States during the events of 2020-2022.", | ||||
| 128 | "format": "HTML", | ||||
| 129 | "hash": "", | ||||
| 130 | "id": "54dccd36-761a-4625-a718-267aeac62478", | ||||
| 131 | "last_modified": null, | ||||
| 132 | "metadata_modified": "2025-10-21T08:59:11.215874", | ||||
| 133 | "mimetype": null, | ||||
| 134 | "mimetype_inner": null, | ||||
| 135 | "name": "Decoding Online Hate in the United States: A BERT-CNN | ||||
| 136 | Analysis of 36 Million Tweets from 2020 to 2022", | ||||
| 137 | "package_id": "2ca1c136-7d06-46e8-8ab7-55b965a55832", | ||||
| 138 | "position": 0, | ||||
| 139 | "resource_type": null, | ||||
| 140 | "size": null, | ||||
| 141 | "state": "active", | ||||
| 142 | "url": | ||||
| 143 | esize=100&sortby=pubdate&citation_for_view=b81TvMMAAAAJ:Se3iqnhoufwC", | ||||
| 144 | "url_type": null | ||||
| 145 | } | ||||
| 146 | ], | ||||
| 102 | "state": "active", | 147 | "state": "active", | ||
| 103 | "tags": [ | 148 | "tags": [ | ||
| 104 | { | 149 | { | ||
| 105 | "display_name": "2024", | 150 | "display_name": "2024", | ||
| 106 | "id": "83ca0aff-0e55-47cc-a3c4-ef8aebf15b19", | 151 | "id": "83ca0aff-0e55-47cc-a3c4-ef8aebf15b19", | ||
| 107 | "name": "2024", | 152 | "name": "2024", | ||
| 108 | "state": "active", | 153 | "state": "active", | ||
| 109 | "vocabulary_id": null | 154 | "vocabulary_id": null | ||
| 110 | }, | 155 | }, | ||
| 111 | { | 156 | { | ||
| 112 | "display_name": "albertogarob", | 157 | "display_name": "albertogarob", | ||
| 113 | "id": "b079b3e9-8dbb-4423-88fb-f095153d314a", | 158 | "id": "b079b3e9-8dbb-4423-88fb-f095153d314a", | ||
| 114 | "name": "albertogarob", | 159 | "name": "albertogarob", | ||
| 115 | "state": "active", | 160 | "state": "active", | ||
| 116 | "vocabulary_id": null | 161 | "vocabulary_id": null | ||
| 117 | }, | 162 | }, | ||
| 118 | { | 163 | { | ||
| 119 | "display_name": "computer-science", | 164 | "display_name": "computer-science", | ||
| 120 | "id": "29cae056-cd7e-43f7-be5b-b25869a3fbf2", | 165 | "id": "29cae056-cd7e-43f7-be5b-b25869a3fbf2", | ||
| 121 | "name": "computer-science", | 166 | "name": "computer-science", | ||
| 122 | "state": "active", | 167 | "state": "active", | ||
| 123 | "vocabulary_id": null | 168 | "vocabulary_id": null | ||
| 124 | }, | 169 | }, | ||
| 125 | { | 170 | { | ||
| 126 | "display_name": "decoding-methods", | 171 | "display_name": "decoding-methods", | ||
| 127 | "id": "8e5cd4be-c8aa-4b2c-906b-89abb0cffb68", | 172 | "id": "8e5cd4be-c8aa-4b2c-906b-89abb0cffb68", | ||
| 128 | "name": "decoding-methods", | 173 | "name": "decoding-methods", | ||
| 129 | "state": "active", | 174 | "state": "active", | ||
| 130 | "vocabulary_id": null | 175 | "vocabulary_id": null | ||
| 131 | } | 176 | } | ||
| 132 | ], | 177 | ], | ||
| 133 | "title": "Decoding Online Hate in the United States: A BERT-CNN | 178 | "title": "Decoding Online Hate in the United States: A BERT-CNN | ||
| 134 | Analysis of 36 Million Tweets from 2020 to 2022", | 179 | Analysis of 36 Million Tweets from 2020 to 2022", | ||
| 135 | "type": "dataset", | 180 | "type": "dataset", | ||
| 136 | "url": | 181 | "url": | ||
| 137 | esize=100&sortby=pubdate&citation_for_view=b81TvMMAAAAJ:Se3iqnhoufwC", | 182 | esize=100&sortby=pubdate&citation_for_view=b81TvMMAAAAJ:Se3iqnhoufwC", | ||
| 138 | "version": null | 183 | "version": null | ||
| 139 | } | 184 | } |
