{"id":13535264,"url":"https://github.com/ianycxu/GCN-with-BERT","last_synced_at":"2025-04-02T00:33:08.941Z","repository":{"id":42053281,"uuid":"183833766","full_name":"ianycxu/GCN-with-BERT","owner":"ianycxu","description":"Graph Convolutional Networks (GCN) with BERT for Coreference Resolution Task [Pytorch][DGL]","archived":false,"fork":false,"pushed_at":"2021-05-18T01:01:57.000Z","size":982,"stargazers_count":141,"open_issues_count":3,"forks_count":27,"subscribers_count":1,"default_branch":"master","last_synced_at":"2024-08-11T16:09:16.857Z","etag":null,"topics":["bert","bert-model","coreference-resolution","gcn","gnn","graph-convolutional-networks","graph-neural-networks","nlp","pytorch"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/ianycxu.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2019-04-28T00:18:25.000Z","updated_at":"2024-07-12T15:27:31.000Z","dependencies_parsed_at":"2022-08-27T15:31:41.988Z","dependency_job_id":null,"html_url":"https://github.com/ianycxu/GCN-with-BERT","commit_stats":null,"previous_names":["ianycxu/rgcn-with-bert"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ianycxu%2FGCN-with-BERT","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ianycxu%2FGCN-with-BERT/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ianycxu%2FGCN-with-BERT/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ianycxu%2FGCN-with-BERT/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ianycxu","download_url":"https://codeload.github.com/ianycxu/GCN-with-BERT/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":246735358,"owners_count":20825222,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["bert","bert-model","coreference-resolution","gcn","gnn","graph-convolutional-networks","graph-neural-networks","nlp","pytorch"],"created_at":"2024-08-01T08:00:52.305Z","updated_at":"2025-04-02T00:33:08.178Z","avatar_url":"https://github.com/ianycxu.png","language":"Jupyter Notebook","funding_links":[],"categories":["BERT  Coreference Resolution","Coreference Resolution:"],"sub_categories":[],"readme":"# Look Again at the Syntax: Relational Graph Convolutional Network for Gendered Ambiguous Pronoun Resolution\n\n## Original Paper\nhttps://www.aclweb.org/anthology/W19-3814/\n\n## Introduction\nWe propose an end-to-end resolver by combining pre-trained BERT with Relational Graph Convolutional Network (R-GCN). R-GCN is used for digesting structural syntactic information and learning better task-specific embeddings. Empirical results demonstrate that, under explicit syntactic supervision and without the need to fine tune BERT, R-GCN's embeddings outperform the original BERT embeddings on the coreference task. Our work obtains the state-of-the-art results on GAP dataset, and significantly improves the snippet-context baseline F1 score from 66.9% to 80.3%. We participated in the 2019 GAP Coreference Shared Task, and our codes are available online. The overall architecture is shown below.\n![](https://i.imgur.com/aAK43SM.png)\n\n\n\n## Dataset we have\nThe data set is Gendered Ambiguous Pronouns (GAP), which is a gender-balanced dataset containing 8908 coreference-labeled pairs sampled from Wikipedia. The dataset contains samples Each sample contains a small paragraph that mentions the potential subject's names later refered by a target pronoun. It also came up with two candidate names for the resolver to choose from. Columns contains: \n\n|  Header        | Description     | \n| :------------- | :----------: |\n|  \u003cstrong\u003eID\u003c/strong\u003e | ID for this sample   | \n|  \u003cstrong\u003eText\u003c/strong\u003e | Text containing pronoun and two names   | \n|  \u003cstrong\u003ePronoun\u003c/strong\u003e | Target pronoun in text   |\n|  \u003cstrong\u003ePronoun-offset\u003c/strong\u003e | Character offset in text   |\n|  \u003cstrong\u003eA\u003c/strong\u003e | Name A in text   |\n|  \u003cstrong\u003eA-offset\u003c/strong\u003e | Position of A in the text |\n|  \u003cstrong\u003eA-coref\u003c/strong\u003e | Whether A confers this pronoun |\n|  \u003cstrong\u003eB\u003c/strong\u003e | Name B in text |\n|  \u003cstrong\u003eB-offset\u003c/strong\u003e | Position of B in the text |\n|  \u003cstrong\u003eA-coref\u003c/strong\u003e | Whether B confers this pronoun |\n\n\n\n## Data Preprocessing\n\nWe use SpaCy as our syntactic denpendency parser. DGL is used to transfer each dependency tree into a graph object. This DGL graph object then can be used as the input for GCN model which is also implemented by DGL. Several graphs are grouped together as a larger DGL batch-graph object for batch training setting.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fianycxu%2FGCN-with-BERT","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fianycxu%2FGCN-with-BERT","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fianycxu%2FGCN-with-BERT/lists"}