{"url":"/method/turl","slug":"turl","name":"TURL","full_name":"TURL: Table Understanding through Representation Learning","full_name_withheld":false,"description_markdown":"Relational tables on the Web store a vast amount of knowledge. Owing to the wealth of such tables, there has been tremendous progress on a variety of tasks in the area of table understanding. However, existing work generally relies on heavily-engineered task- specific features and model architectures. In this paper, we present TURL, a novel framework that introduces the pre-training/fine- tuning paradigm to relational Web tables. During pre-training, our framework learns deep contextualized representations on relational tables in an unsupervised manner. Its universal model design with pre-trained representations can be applied to a wide range of tasks with minimal task-specific fine-tuning.\r\nSpecifically, we propose a structure-aware Transformer encoder to model the row-column structure of relational tables, and present a new Masked Entity Recovery (MER) objective for pre-training to capture the semantics and knowledge in large-scale unlabeled data. We systematically evaluate TURL with a benchmark consisting of 6 different tasks for table understanding (e.g., relation extraction, cell filling). We show that TURL generalizes well to all tasks and substantially outperforms existing methods in almost all instances.","description_state":"present","introduced_year":null,"introduced_by":{"title":"TURL: Table Understanding through Representation Learning","paper":"/paper/turl-table-understanding-through","first_author":"Xiang Deng","n_authors":5,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/turl-table-understanding-through"},"source":{"url":"https://arxiv.org/abs/2006.14806v2","title":"TURL: Table Understanding through Representation Learning","url_on_a_paper_host":true},"code_snippet_url":null,"code_snippet_url_on_a_code_host":false,"categories":[{"area":"General","area_id":"general","collection":"Deep Tabular Learning","url":"/methods/category/deep-tabular-learning","pwc_aliases":[]}],"n_papers_tagged":3,"archive_num_papers":3,"papers_newest_first":[{"paper":null,"title":"Evaluating LLMs on Entity Disambiguation in Tables","date":"2024-08-12","arxiv_id":"2408.06423","n_code_links":0,"syntology":null},{"paper":"/paper/sotab-the-wdc-schema-org-table-annotation","title":"SOTAB: The WDC Schema.org Table Annotation Benchmark","date":"2023-01-09","arxiv_id":null,"n_code_links":1,"syntology":null},{"paper":"/paper/turl-table-understanding-through","title":"TURL: Table Understanding through Representation Learning","date":"2020-06-26","arxiv_id":"2006.14806","n_code_links":1,"syntology":{"ran":1,"of":1,"unverified":0,"pointer_only":0}}],"papers_shown":3,"tasks":[{"task":"/task/table-annotation","name":"Table annotation","papers":3},{"task":"/task/column-type-annotation","name":"Column Type Annotation","papers":2},{"task":"/task/columns-property-annotation","name":"Columns Property Annotation","papers":2},{"task":"/task/cell-entity-annotation","name":"Cell Entity Annotation","papers":1},{"task":"/task/data-integration","name":"Data Integration","papers":1},{"task":"/task/decoder","name":"Decoder","papers":1},{"task":"/task/entity-disambiguation","name":"Entity Disambiguation","papers":1},{"task":"/task/missing-values","name":"Missing Values","papers":1},{"task":"/task/relation-extraction","name":"Relation Extraction","papers":1},{"task":"/task/representation-learning","name":"Representation Learning","papers":1}],"tasks_shown":10,"n_tasks":10,"usage_by_year":[{"year":"2020","papers":1},{"year":"2023","papers":1},{"year":"2024","papers":1}],"row_source":"methods_table","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/turl"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}