2nd Workshop for Natural Language Processing Open Source Software (NLP-OSS)
19 Nov 2020 @ EMNLP 2020 (Virtual Workshop)
With great scientific breakthrough comes solid engineering and open communities. The Natural Language Processing (NLP) community has benefited greatly from the open culture in sharing knowledge, data, and software. The primary objective of this workshop is to further the sharing of insights on the engineering and community aspects of creating, developing, and maintaining NLP open source software (OSS), which we seldom talk about in scientific publications. Our secondary goal is to promote synergies between different open source projects and encourage cross-software collaborations and comparisons.
We refer to Natural Language Processing OSS as an umbrella term that not only covers traditional syntactic, semantic, phonetic, and pragmatic applications; we extend the definition to include task-specific applications (e.g., machine translation, information retrieval, question-answering systems), low-level string processing that contains valid linguistic information (e.g. Unicode creation for new languages, language-based character set definitions) and machine learning/artificial intelligence frameworks with functionalities focusing on text applications.
In the earlier days of NLP, linguistic software was often monolithic and the learning curve to install, use, and extend the tools was steep and frustrating. More often than not, NLP OSS developers/users interact in siloed communities within the ecologies of their respective projects. In addition to the engineering aspects of NLP software, the open source movement has brought a community aspect that we often overlook in building impactful NLP technologies.
An example of precious OSS knowledge comes from SpaCy developer Montani (2017), who shared her thoughts and challenges of maintaining commercial NLP-OSS, such as handling open issues on the issue tracker, model release and packaging strategy and monetizing NLP OSS for sustainability.
More recently, the Transformers library created by Hugging Face, has gathered much interest from the community by open sourcing implementations to use pretrained weights of BERT-like models, in a clean and well-organized structure. The interoperability of various pretrained models trained with different tools in one library enables quick benchmarking across the models, as well as developing best practices for reading/saving serialized interoperable models.
We hope that the NLP-OSS workshop becomes the intellectual forum to collate various open source knowledge beyond the scientific contribution, announce new software/features, promote the open source culture and best practices that go beyond the conferences.
Call for Papers
We invite full papers (8 pages) or short papers (4 pages) on topics related to NLP-OSS broadly categorized into (i) software development, (ii) scientific contribution and (iii) NLP-OSS case studies.
- Software Development
- Designing and developing NLP-OSS
- Licensing issues in NLP-OSS
- Backward compatibility and stale code in NLP-OSS
- Growing, maintaining and motivating an NLP-OSS community
- Best practices for NLP-OSS documentation and testing
- Contribution to NLP-OSS without coding
- Incentivizing OSS contributions in NLP
- Commercialization and Intellectual Property of NLP-OSS
- Defining and managing NLP-OSS project scope
- Issues in API design for NLP
- NLP-OSS software interoperability
- Analysis of the NLP-OSS community
- Scientific Contribution
- Surveying OSS for specific NLP task(s)
- Demonstration, introductions and/or tutorial of NLP-OSS
- Small but useful NLP-OSS
- NLP components in ML OSS
- Citations and references for NLP-OSS
- OSS and experiment replicability
- Gaps between existing NLP-OSS
- Task-generic vs task-specific software
- Case studies
- Case studies of how a specific bug is fixed or feature is added
- Writing wrappers for other NLP-OSS
- Writing open-source APIs for open data
- Teaching NLP with OSS
- NLP-OSS in the industry
Authors are invited to submit a
- Full paper up to 8 pages of content or
- Short paper up to 4 pages of content
Submissions can be non-archival and be presented in the NLP-OSS workshop, but we would still require at least a 4-page submission so that reviewers have enough information to make the acceptance/rejection decision. This non-archival option is helpful for author(s) who wants to publish or had published the work elsewhere and would like to present/discuss pertinent NLP-OSS related work to the workshop PCs and attendees.
All papers are allowed unlimited but sensible pages for references. Final camera-ready versions will be allowed an additional page of content to address reviewers’ comments.
Due to the nature of open source software, we find it a bit tricky to “anonymize” “open source”. For this reason, we don’t require your publication to be anonymous. However, if you prefer your paper to be anonymized, please mask any identifiable phrase with REDACTED. We have an option setup in softconf so that you can explicitly opt-in / opt-out of anonymity.
Submission should be formatted according to the EMNLP 2020 LaTeX or MS Word templates at https://2020.emnlp.org/files/emnlp2020-templates.zip.
Submissions should be uploaded to Softconf conference management system at https://www.softconf.com/emnlp2020/nlposs/.
Note: Paper can be dual-submitted to both EMNLP 2020 and the NLP-OSS workshop.
The 2nd NLP-OSS workshop will be co-located with the EMNLP 2020 conference.
- Paper submission:
05 August 202019 August 2020
- Paper Reviews Starts:
07 August 202020 August 2020
- Paper Reviews Due:
07 September 202020 September 2020
- Notification of Acceptance:
10 September 2020 22 September 202025 September 2020
- Camera-Ready Version:
05 October 2020 10 October 202009 Oct 2020
- Workshop: 19 November 2020
Principles of Good Machine Learning Systems Design
Chip Huyen, Snorkel AI / Stanford University
This talk covers what it means to operationalize Machine Learning (ML) models. It starts by analyzing the difference between ML in research vs. in production, ML systems vs. traditional software, as well as myths about ML production. It then goes over the principles of good ML systems design and introduces an iterative framework for ML systems design, from scoping the project, data management, model development, deployment, maintenance, to business analysis. It covers the differences between DataOps, ML Engineering, MLOps, and data science, and where each fits into the framework. The talk ends with a survey of the ML production ecosystem, the economics of open source, and open-core businesses.
Chip Huyen is an engineer who develops tools and best practices for machine learning production. She’s currently with Snorkel AI and she’ll be teaching Machine Learning Systems Design at Stanford. Previously, she was with Netflix, NVIDIA, Primer. She’s also the author of four bestselling Vietnamese books.
On Typing: Historical and Potential Interactions in Word-processing
Spencer Kelly, Freelance Developer
People love typing, in a surprising and universal way. In this talk we look at the development of word-processing, and the design-decisions in this historic interface. Can NLP contribute to word-processing, without making it worse? What would a text-centered computer really look like? We look at the history of punctuation, keyboards, and markup languages. We look at Wikipedia, text-editors, and data structures - with the goal of authoring usable data in text.
Spencer Kelly is the author of compromise, - a small natural language processing library for the browser. He is a web developer, and maintainer of open-source libraries. His background is in the semantic web and Wikipedia. Today his work focuses on creating infographics. His open-source work is funded by freelance web development. He is from Toronto, Canada.
An Introduction to Transfer Learning in NLP and HuggingFace
Thomas Wolf, Huggingface
In this talk I’ll start by introducing the recent breakthroughs in NLP that resulted from the combination of Transfer Learning schemes and Transformer architectures. The second part of the talk will be dedicated to an introduction of the open-source tools released by HuggingFace, in particular our Transformers, Tokenizers and Datasets libraries and our models.
Thomas Wolf is co-founder and Chief Science Officer of HuggingFace. His team is on a mission to catalyze and democratize NLP research. Prior to HuggingFace, Thomas gained a Ph.D. in physics, and later a law degree. He worked as a physics researcher and a European Patent Attorney.
The timezone for the program schedule below are in Pacific Time (Los Angeles).
0500 - 0530 Opening Remarks
0530 - 0630 Invited Talk 1: Spencer Kelly (YouTube)
0630 - 0730 Invited Talk 2: Thomas Wolf (YouTube)
0730 - 0900 Talks Session 1 (Watch on your own, authors encouraged to mend their slidelive chat but not complusory)
A Framework to Assist Chat Operators of Mental Healthcare Services
Thiago Madeira, Heder Bernardino, Jairo Francisco De Souza, Henrique Gomide, Nathália Munck Machado, Bruno Marcos Pinheiro da Silva and Alexandre Vieira Pereira Pacelli
ARBML: Democritizing Arabic Natural Language Processing Tools
Zaid Alyafeai and Maged Al-Shaibani
CLEVR Parser: A Graph Parser Library for Geometric Learning on Language Grounded Image Scenes
Raeid Saqur and Ameet Deshpande
End-to-end NLP Pipelines in Rust
Fair Embedding Engine: A Library for Analyzing and Mitigating Gender Bias in Word Embeddings
Vaibhav Kumar, Tenzin Bhotia and Vaibhav Kumar
Flexible retrieval with NMSLIB and FlexNeuART
Leonid Boytsov and Eric Nyberg
fugashi, a Tool for Tokenizing Japanese in Python
Going Beyond T-SNE: Exposing whatlies in Text Embeddings
Vincent Warmerdam, Thomas Kober and Rachael Tatman
Howl: A Deployed, Open-Source Wake Word Detection System
Raphael Tang, Jaejun Lee, Afsaneh Razi, Julia Cambre, Ian Bicking, Jofish Kaye and Jimmy Lin
iNLTK: Natural Language Toolkit for Indic Languages
0900 - 1430 Long Break
1430 - 1600 Talks Session 2 (Watch on your own, authors encouraged to mend their slidelive chat but not complusory)
iobes: A Library for Span-Level Processing
jiant: A Software Toolkit for Research on General-Purpose Text Understanding Models
Yada Pruksachatkun, Phil Yeres, Haokun Liu, Jason Phang, Phu Mon Htut, Alex Wang, Ian Tenney and Samuel R. Bowman
KLPT -- Kurdish Language Processing Toolkit
Open Korean Corpora: A Practical Report
Won Ik Cho, Sangwhan Moon and Youngsook Song
Open-Source Morphology for Endangered Mordvinic Languages
Jack Rueter, Mika Hämäläinen and Niko Partanen
Pimlico: A toolkit for corpus-processing pipelines and reproducible experiments
PySBD: Pragmatic Sentence Boundary Disambiguation
Nipun Sadvilkar and Mark Neumann
SacreROUGE: An Open-Source Library for Using and Developing Summarization Evaluation Metrics
Daniel Deutsch and Dan Roth
TextAttack: Lessons learned in designing Python frameworks for NLP
John Morris, Jin Yong Yoo and Yanjun Qi
TOMODAPI: A Topic Modeling API to Train, Use and Compare Topic Models
Pasquale Lisena, Ismail Harrando, Oussama Kandakji and Raphael Troncy
User-centered & Robust NLP OSS: Lessons Learned from Developing & Maintaining RSMTool
Nitin Madnani and Anastassia Loukina
WAFFLE: A Graph for WordNet Applied to FreeForm Linguistic Exploration
Berk Ekmekci and Blake Howald
1600 - 1730 Gather-town (Live) + Poster QnA
1730 - 1800 Short Break
1800 - 1900 Invited Talk: Chip Huyen (YouTube)
1900 - 1930 Closing Remarks
Lucy Park, Upstage
Lucy is a NLP/ML engineer at Upstage. She has participated in some open source projects, particularly KoNLPy which is a tool for Korean NLP, and is also interested in open data. She received her Ph.D. in Data Mining from Seoul National University in 2016, where she has pursued various studies on text mining in the fields of manufacturing, political science, and multimedia. After her studies, she joined NAVER, a South Korea based search-engine company and worked on machine translation.
Masato Hagiwara, Octanove Labs LLC
Masato Hagiwara is an independent NLP/ML engineer and researcher at Octanove Labs. He received his Ph.D. degree in Information Science from Nagoya University in 2009. During his Ph.D., he worked at Google and Microsoft Research as an intern, and thereafter at Baidu Japan and Rakuten Institute of Technology, focusing on search engine-related language processing research. Most recently he was working as a Senior Machine Learning Engineer at Duolingo, focusing on educational applications of NLP. He received several paper awards from Japanese domestic conferences for his work on knowledge acquisition and transliteration. He also co-organized several workshops and shared tasks, including NLP-OSS 2018.
Dmitrijs Milajevs, KPMG LLP.
Dmitrijs Milajevs is a data scientist at KMPG. Previously, he evaluated information retrieval systems at National Institute of Standards and Technology (NIST). He has defended a Ph.D. thesis on evaluation of compositional models in distributional semantics.
Nelson Liu, Stanford University
Nelson Liu is a computer science Ph.D. student at Stanford, where he works in the Stanford NLP Group. He has contributed to the AllenNLP, torchtext, and scikit-learn projects at various points in time.
Geeticka Chauhan, Massachusetts Institute of Technology
Geeticka Chauhan is a Ph.D. student at MIT, working on NLP for healthcare advised by Prof. Peter Szolovits. Her master thesis focused on revealing the reproducibility and generalizability problems in Relation Extraction, and experimentally showed the importance of streamlining evaluation methods in NLP challenges.
Liling Tan, Amazon
Liling is a machine learning scientist at Amazon building machine translation and language technologies. Previously, he is a research scientist at Rakuten Institute of Technology and before that an early stage researcher at Saarland University. He has been actively involved in corpora creation/maintenance, Asian NLP and machine translation. He co-organized the Workshop on NLP for Similar Languages, Varieties and Dialects (VarDial 2014-16).
- Aline Paes, Universidade Federal Fluminense
- Amandalynne Paullada, University of Washington
- Amittai Axelrod, DiDi Chuxing (Los Angeles)
- Anca Dumitrache, FD Mediagroep
- Arwen Twinkle Griffioen, Zendesk Inc.
- Carolina Scarton, University of Sheffield
- Chris Hokamp, AYLIEN
- Christian Federmann, Microsoft Research
- Dan Simonson, BlackBoiler, LLC
- Daniel Braun, TU Muchen
- Dave Howcroft, Heriot-Watt University
- David Przybilla, Idio
- Delip Rao, AI Foundation
- Denny Britz, Prediction Machines
- Ehsan Khoddammohammadi, Elsiever
- Eleftherios Avramidis, German Research Center for Artificial Intelligence
- Elijah Rippeth, MITRE Corporation
- Emiel van Miltenburg, Vrije Universiteit Amsterdam
- Emily Dinan, Facebook AI
- Eric Schles, New York University / Sema4
- Fabio Kepler, Unbabel
- Francis Bond, Nanyang Technological University
- Fred Blain, University of Sheffield
- Gerard Dupont, Airbus
- Ian Soboroff, NIST
- Ignatius Ezeani, Lancaster Uni
- Ines Montani, Explosion AI
- James Bradbury, Google
- Joel Nothman, University of Sydney
- Karin Sim Smith, Lingo24
- Kevin Cohen, University of Colorado Boulder
- KhengHui Yeo, Institute for Infocomm Research
- Laura Martinus, Explore AI
- Madison May, Indico Data Solutions
- Marcel Bollmann, University of Copenhagen
- Marcos Zampieri, University of Wolverhampton
- Mary Ellen Foster, University of Glasgow
- Marzieh Fadaee, University of Amsterdam
- Matthew Honnibal, Explosion AI
- Micah Shlain, Allen Institute for Artificial Intelligence
- Michael Wayne Goodman, Nanyang Technological University
- Mohd Sanad Zaki Rizvi, Microsoft Research India
- Moshe Wasserblat, Intel
- Muthu Kumar Chandrasekaran, NUS, SG
- Nahid Alam, Ople Inc
- Paul P Liang, Carnegie Mellon University
- Philipp Koehn, Johns Hopkins University
- Sandya Mannarswamy , Independent Researcher
- Shamil Chollampatt, Rakuten Institute of Technology
- Sharat Chikkerur, Microsoft
- Shilpa Suresh, Singapore Managment University
- Shubhanshu Mishra, Twitter
- Steve DeNeefe, SDL Research Labs
- Steve Sloto, AWS AI
- Steven Bethard, University of Arizona
- Steven Bird, Charles Darwin University
- Sung Kim, NAVER Corp.
- Svitlana Vakulenko, University of Amsterdam
- Tareq Al-Moslmi, University of Bergen
- Thomas Kober, Rasa Technologies GmbH
- Tilahun Abedissa, Addis Ababa University
- Tommaso Teofili, Roma Tre University / Red Hat
- Tommi A Pirinen, University of Hamburg
- Varun Kumar, Amazon Alexa
- Vlad Niculae, Instituto de Telecomunicações
- Yves Peirsman, NLP Town