{"url":"/dataset/uml-classes-with-specs","name":"UML Classes With Specs","full_name":"UML Class Diagrams Paired With Their English Specifications","description_markdown":"# Repository for UML-English data\r\nThis repository contains the data used for \"Extraction of UML Class Diagrams from Natural Language Specification\" (Yang et al. 2022)\r\n\r\n## Getting the dataset\r\nTo get the entire dataset, you must download the release containing `dataset.tar.gz`.\r\n\r\n## Structure of the dataset\r\n\r\n* `dataset.tar.gz`: archive that contains all the files\r\n* `fragments.csv`: file that lists UML fragments and their characteristics\r\n* `labels.csv`: file that contains the labels received in the crowdsourcing effort\r\n* `models.csv`: file that lists UML class diagrams and their characteristics\r\n* `zoo/`: folder that contains all the UML data itself, such as pictures and UML encodings\r\n\r\n## Making use of the dataset\r\nUnzip the tarball first.\r\n\r\n### Opening the image of a certain UML model\r\nOpen `models.csv` to read the list of available models. Copy its name and search in the `zoo/` folder for `.png` files starting with that name. For example, the ACME model has an image in the `zoo/` folder called `ACME.png`.\r\n\r\n```bash\r\nls zoo/ACME.png\r\ncode zoo/ACME.png # any other image visualizer\r\n```\r\n\r\n### Opening the image of a certain fragment\r\nFragment files are named in the following pattern.\r\n\r\nClass fragments:\r\n```\r\n(ModelName)_(class)(number).png\r\n```\r\n\r\nRelationship fragments:\r\n```\r\n(ModelName)_(rel)(number).png\r\n```\r\n\r\nSimilarly, you can visualize them.\r\n```bash\r\ncode zoo/CFG_class0.png\r\n```\r\n\r\n### Finding the image of a fragment starting from a label\r\n1. Browse through `labels.csv` and find the line that has the label of interest.\r\n2. Every label has a `fragment_id`, which can be indexed in `fragments.csv`. Find the ID for the label of interest.\r\n3. Inside `fragments.csv`, search for the line where the column value of `unique_id` equals `fragment_id` from Step 2.\r\n4. Proceed like in the previous [section](#opening-the-image-of-a-certain-fragment)","description_withheld":null,"homepage":"https://github.com/XsongyangX/uml-classes-and-specs","introduced_date":"2022-10-26","introduced_date_note":null,"introduced_by":{"paper":"/paper/towards-automatically-extracting-uml-class","title":"Towards Automatically Extracting UML Class Diagrams from Natural Language Specifications","first_author":"Song Yang","url":null},"license":null,"modalities":[],"tasks":[{"name":"Model extraction","url":"/task/model-extraction","datasets_with_task":"/datasets/task/model-extraction"}],"languages":[{"name":"English","url":"/datasets/language/english"}],"variants":["UML Classes With Specs"],"data_loaders":[],"num_papers_in_archive":1,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/model-extraction-on-uml-classes-with-specs","task":"Model extraction","dataset_variant":"UML Classes With Specs","rows":1,"metrics":["Exact Match"],"first_row_in_archive_order":{"model":"three-step-original","paper":"/paper/towards-automatically-extracting-uml-class","metrics":{"Exact Match":"0.171"},"code_links":[{"title":"XsongyangX/uml-translation-3step","url":"https://github.com/XsongyangX/uml-translation-3step"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/towards-automatically-extracting-uml-class","title":"Towards Automatically Extracting UML Class Diagrams from Natural Language Specifications","date":"2022-10-26","rows_on_this_dataset":1,"code_links":1,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":0,"samples_harvested":0,"samples_ran":0,"samples_unverified":0,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}