{"id":18835613,"url":"https://github.com/nschneid/arabic-tagger","last_synced_at":"2025-07-13T17:34:33.574Z","repository":{"id":3301693,"uuid":"4343396","full_name":"nschneid/arabic-tagger","owner":"nschneid","description":"AQMAR Arabic Tagger: Sequence tagger with cost-augmented structured perceptron training","archived":false,"fork":false,"pushed_at":"2013-08-28T04:07:27.000Z","size":9153,"stargazers_count":42,"open_issues_count":1,"forks_count":18,"subscribers_count":6,"default_branch":"master","last_synced_at":"2025-04-14T05:39:22.387Z","etag":null,"topics":["arabic","arabic-language","arabic-nlp","arabic-wikipedia","java","named-entities","nlp","nlp-machine-learning","sequence-tagger","tagger"],"latest_commit_sha":null,"homepage":null,"language":"Java","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"gpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/nschneid.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2012-05-16T04:07:50.000Z","updated_at":"2024-12-09T10:50:44.000Z","dependencies_parsed_at":"2022-08-26T12:33:16.112Z","dependency_job_id":null,"html_url":"https://github.com/nschneid/arabic-tagger","commit_stats":null,"previous_names":[],"tags_count":1,"template":false,"template_full_name":null,"purl":"pkg:github/nschneid/arabic-tagger","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nschneid%2Farabic-tagger","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nschneid%2Farabic-tagger/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nschneid%2Farabic-tagger/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nschneid%2Farabic-tagger/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/nschneid","download_url":"https://codeload.github.com/nschneid/arabic-tagger/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nschneid%2Farabic-tagger/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":265179233,"owners_count":23723430,"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":["arabic","arabic-language","arabic-nlp","arabic-wikipedia","java","named-entities","nlp","nlp-machine-learning","sequence-tagger","tagger"],"created_at":"2024-11-08T02:16:41.935Z","updated_at":"2025-07-13T17:34:33.522Z","avatar_url":"https://github.com/nschneid.png","language":"Java","funding_links":[],"categories":[],"sub_categories":[],"readme":"AQMAR Arabic Tagger\n===================\n\nThis package provides a sequence tagger implementation customized for Arabic features, \nincluding a named entity detection model especially intended for Arabic Wikipedia. \nIt was trained on labeled ACE and ANER data as well as an unlabeled Wikipedia corpus.\nLearning is with the structured perceptron, optionally in a cost-augmented fashion.\nFeature extraction is handled as a preprocessing step prior to learning/decoding.\n\nThe tagger was used for the experiments reported in\n\n - Behrang Mohit, Nathan Schneider, Rishav Bhowmick, Kemal Oflazer, and Noah A. Smith (2012),\n  Recall-Oriented Learning of Named Entities in Arabic Wikipedia. _Proceedings of EACL_.\n\nand accompanies the AQMAR Arabic Wikipedia Named Entity Corpus also described in that work; \nboth can be obtained at\n\n  http://www.ark.cs.cmu.edu/AQMAR/\n\nThe Java tagger was adapted from Michael Heilman's supersense tagger implementation \nfor English (http://www.ark.cs.cmu.edu/mheilman/questions/). It requires a minimum \nJava version of 1.6. Feature extraction uses Python and depends on the MADA toolkit \n(http://www1.ccls.columbia.edu/MADA/; version 3.1 was used for the Named Entity Corpus).\n\nThe AQMAR Arabic Tagger is released under the GNU General Public License (GPL) version 3 or \nlater; see LICENSE. (Michael Heilman's supersense tagger, which we modify, was originally \nreleased in 2011 under GPL version 2 or later; the JSAP library, which we link to, was \noriginally released by Martian Software in 2011 under the Lesser GNU Public License.)\n\n\nContents\n--------\n\n- eval/\n  \n  README and scripts for NER evaluation.\n  \n- featExtract/\n  \n  README and scripts for feature extraction.\n  \n- lib/\n  \n  External libraries required for the Java tagger.\n  \n- model/\n  \n  Serialized tagging models, namely the best Arabic Wikipedia tagger reported in the EACL paper.\n\n- src/\n  \n  Java source files for the tagger.\n  \n- arabic-tagger.jar\n  \n  Compiled Java program for training and decoding with the tagger.\n\n- build.sh\n  \n  Script for compiling the Java sources.\n  \n- sample.properties\n  \n  An example properties file that can be used to specify options for the tagger. \n  Options may alternatively be passed as command-line flags; if an option is specified \n  in both places, the command-line value will take precedence.\n\n- LICENSE\n- README\n- VERSION\n\n\nUsage\n-----\n\nExtracting features for text data: See featExtract/README.txt\n\n### Running the Arabic named entity tagger\n\nFor example, the following command will use the existing named entity model in the model/ directory:\n\n    java -Xmx8000m -XX:+UseCompressedOops -jar arabic-tagger.jar \n  \t\t--load model/arabic-ner-superROP200.selfROP100.ser.gz \n\t\t--test-predict featExtract/sample.bio.nerFeats --usePrevLabel true\n\t\t--properties sample.properties \u003e predictions.out\n\n### Training a tagging model\n\nHere is an example command for training a model on the sample feature-extracted data:\n\n    java -Xmx8000m -XX:+UseCompressedOops -jar arabic-tagger.jar \n  \t\t--save model/sample-model.ser.gz --iters 10 --no-averaging\n\t\t--labels featExtract/sample.labels --train featExtract/sample.nerFeats --debug --disk --weights\n\t\t--properties sample.properties \u003e weights.out\n\nor boundaries only:\n\n    java -Xmx8000m -XX:+UseCompressedOops -jar arabic-tagger.jar \n  \t\t--save model/sample-model.ser.gz --iters 10 --no-averaging\n\t\t--labels featExtract/bio.labels --train featExtract/sample.bio.nerFeats --debug --disk --weights\n\t\t--properties sample.properties \u003e weights.out\n\n__Until [this bug](https://github.com/nschneid/arabic-tagger/issues/1) is fixed, we recommend specifying `--no-averaging` for training.__\n\nFor details about options, run\n  \n    java -jar arabic-tagger.jar --help\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnschneid%2Farabic-tagger","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fnschneid%2Farabic-tagger","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnschneid%2Farabic-tagger/lists"}