{"id":22458876,"url":"https://github.com/s1998/re-label-noise","last_synced_at":"2025-03-27T14:17:09.575Z","repository":{"id":67610949,"uuid":"189835305","full_name":"s1998/re-label-noise","owner":"s1998","description":"Code for modelling label noie in distantly supervised relation extraction","archived":false,"fork":false,"pushed_at":"2022-10-11T17:23:38.000Z","size":2614,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-02-01T18:27:39.434Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"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/s1998.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,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2019-06-02T10:56:14.000Z","updated_at":"2022-10-11T17:10:02.000Z","dependencies_parsed_at":null,"dependency_job_id":"014b45bb-bbf9-4a44-b691-8ded27c9f0ca","html_url":"https://github.com/s1998/re-label-noise","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/s1998%2Fre-label-noise","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/s1998%2Fre-label-noise/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/s1998%2Fre-label-noise/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/s1998%2Fre-label-noise/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/s1998","download_url":"https://codeload.github.com/s1998/re-label-noise/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":245858881,"owners_count":20684062,"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":[],"created_at":"2024-12-06T08:42:06.258Z","updated_at":"2025-03-27T14:17:09.557Z","avatar_url":"https://github.com/s1998.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Relation extraction (modelling the label noise)\nCode for modelling label noie in distantly supervised relation extraction.\n\n# Usage help\n\n```\nusage: main.py [-h] [--encoder ENCODER] [--selector SELECTOR]\n               [--loss_type LOSS_TYPE] [--bs_val BS_VAL] [--dataset DATASET]\n               [--chkpt_pt CHKPT_PT] [--l2_val L2_VAL]\n\nRunning code for relation extraction.\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --encoder ENCODER     select the encoders from pcnn ,pbrnn ,pcnn2 (stacked\n                        pcnn) ,rnn ,brnn ,crnn ,crnn2 ,bgwa\n  --selector SELECTOR   select the bag selector from att, cross_sent_max\n  --loss_type LOSS_TYPE\n                        select the loss type from none, extra (layer for noise\n                        modelling), hard (bootstrapping), soft (bootstrapping)\n  --bs_val BS_VAL       select the bootstrapping value (only valid if loss is\n                        of type hard/soft)\n  --dataset DATASET     select the dataset from nyt/wiki\n  --chkpt_pt CHKPT_PT   path to saved model (empty if no checkpoint to load\n                        from)\n  --l2_val L2_VAL       l2 lambda value\n```\n# Results\n\nResults on NYT dataset\n\n## Results for cross senence maxpooling on NYT dataset\n\n```\nEncoder  | Selector       |  AUC\n---------------------------------\npcnn     | att            | 0.338\nrnn      | att            | 0.333\nbrnn     | att            | 0.344\npcnn     | cross-sent-max | 0.369\nrnn      | cross-sent-max | 0.385\nbrnn     | cross-sent-max | 0.383\n```\n\n## Results for modelling label noise on NYT dataset\n\n```\nMechanism                  | AUC\n----------------------------------\nPCNN + ATT                 | 0.338\nPCNN + ATT + extra_layer   | 0.348\n```\n\nWe found that value of AUC score decreased for bootsrapping methods both hard and soft.\n\n\n# References \n1. [THUNLP's relation extraction](https://github.com/thunlp/OpenNRE)\n2. [Adversarial methods for relation extraction](https://github.com/jxwuyi/AtNRE)\n3. [Soft label methods fro relation extraction](https://github.com/tyliupku/soft-label-RE)\n\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fs1998%2Fre-label-noise","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fs1998%2Fre-label-noise","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fs1998%2Fre-label-noise/lists"}