{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/cloudscan-a-configuration-free-invoice","title":"CloudScan - A configuration-free invoice analysis system using recurrent neural networks","arxiv_id":"1708.07403","date":"2017-08-24","proceeding":null,"authors":["Rasmus Berg Palm","Ole Winther","Florian Laws"],"abstract":"We present CloudScan; an invoice analysis system that requires zero\nconfiguration or upfront annotation. In contrast to previous work, CloudScan\ndoes not rely on templates of invoice layout, instead it learns a single global\nmodel of invoices that naturally generalizes to unseen invoice layouts. The\nmodel is trained using data automatically extracted from end-user provided\nfeedback. This automatic training data extraction removes the requirement for\nusers to annotate the data precisely. We describe a recurrent neural network\nmodel that can capture long range context and compare it to a baseline logistic\nregression model corresponding to the current CloudScan production system. We\ntrain and evaluate the system on 8 important fields using a dataset of 326,471\ninvoices. The recurrent neural network and baseline model achieve 0.891 and\n0.887 average F1 scores respectively on seen invoice layouts. For the harder\ntask of unseen invoice layouts, the recurrent neural network model outperforms\nthe baseline with 0.840 average F1 compared to 0.788.","url_abs":"http://arxiv.org/abs/1708.07403v1","url_pdf":"http://arxiv.org/pdf/1708.07403v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"cloudscan-a-configuration-free-invoice","repo_url":"https://github.com/naiveHobo/InvoiceNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.07403","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}