- Corpora - ELRA lists

PhD position on efficient E2E models for Automatic Speech Recognition
by Daniele Falavigna 07 Aug '23

07 Aug '23

--- Apologies for multiple posting --- The Speech Technology group (SpeechTek) at Fondazione Bruno Kessler <https://www.fbk.eu/en/> (Trento, Italy) in conjunction with the ICT International Doctorate School of the University of Trento <https://iecs.unitn.it/> is pleased to announce the availability of the following fully-funded PhD position: *TITLE*: Efficient E2E models for automatic speech recognition in multi-speaker scenarios *DESCRIPTION*: In spite of the recent progress in speech technologies, processing and understanding conversational spontaneous speech is still an open issue, in particular in presence of challenging acoustic conditions as those posed by dinner party scenarios. Although enormous progresses have been made recently in a variety of speech processing tasks (such as speech enhancement, speech separation, speech recognition, spoken language understanding), targeting also multi-speaker speech recognition, a unified established solution is still far from being available. Moreover, the computational complexity of current approaches is extremely high, making an actual deployment in low-end or IoT devices not feasible in practice. The candidate will advance the current state-of-the-art in speech processing (in particular for separation, enhancement and recognition) towards developing a unified solution, possibly based on self-supervised or unsupervised approaches, for automatic speech recognition in dinner party scenarios, as those considered in the CHiME challenges ( https://arxiv.org/abs/2306.13734). *CONTACTS*: brutti(a)fbk.eu <guerini(a)fbk.eu> *COMPLETE DETAILS AVAILABLE AT*: https://iecs.unitn.it/education/admission/reserved-topic-scholarships#A9 ------------------------------------------------------------------------------------------------ Giuseppe Daniele Falavigna Fondazione Bruno Kessler Via Sommarive 18 - 38123 Povo - Trento, Italy mail:falavi@fbk.eu - tel:+39(0)461314562 - fax:+39(0)461314591 HomePage: https://speechtek.fbk.eu/people/profile/falavi ------------------------------------------------------------------------------------------------- -- -- Le informazioni contenute nella presente comunicazione sono di natura privata e come tali sono da considerarsi riservate ed indirizzate esclusivamente ai destinatari indicati e per le finalità strettamente legate al relativo contenuto. Se avete ricevuto questo messaggio per errore, vi preghiamo di eliminarlo e di inviare una comunicazione all’indirizzo e-mail del mittente. -- The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. If you received this in error, please contact the sender and delete the material.

1 0

Shared Task: Prompting Large Language Models as Explainable Metrics
by Juri Opitz 07 Aug '23

07 Aug '23

Dear colleagues, you are invited to participate in the Eval4NLP 2023 shared task on **Prompting Large Language Models as Explainable Metrics**. Please find more information below and on the shared task webpage: https://eval4nlp.github.io/2023/shared-task.html Important Dates - Shared task announcement: August 02, 2023 - Dev phase: August 07, 2023 - Test phase: September 18, 2023 - System Submission Deadline: September 23, 2023 - System paper submission deadline: October 5, 2023 - System paper camera ready submission deadline: October 12, 2023 All deadlines are 11.59 pm UTC -12h (“Anywhere on Earth”). The timeframe of the test phase may change. Please regularly check the shared task webpage: https://eval4nlp.github.io/2023/shared-task.html. ** Overview ** With groundbreaking innovations in unsupervised learning and scalable architectures the opportunities (but also risks) of automatically generating audio, images, video and text, seem overwhelming. Human evaluations of this content are costly and are often infeasible to collect. Thus, the need for automatic metrics that reliably judge the quality of generation systems and their outputs, is stronger than ever. Current state-of-the-art metrics for natural language generation (NLG) still do not match the performance of human experts. They are mostly based on black-box language models and usually return a single quality score (sentence-level), making it difficult to explain their internal decision process and their outputs. The release of APIs to large language models (LLMs), like ChatGPT and the recent open-source availability of LLMs like LLaMA has led to a boost of research in NLP, including LLM-based metrics. Metrics like GEMBA [*] explore the prompting of ChatGPT and GPT4 to directly leverage them as metrics. Instructscore [*] goes in a different direction and finetunes a LLaMA model to predict a fine grained error diagnosis of machine translated content. We notice that current work (1) does not systematically evaluate the vast amount of possible prompts and prompting techniques for metric usage, including, for example, approaches that explain a task to a model or let the model explain a task itself, and (2) rarely evaluates the performance of recent open-source LLMs, while their usage is incredibly important to improve the reproducibility of metric research, compared to closed-source metrics. This year’s Eval4NLP shared task, combines these two aspects. We provide a selection of open-source, pre-trained LLMs. The task is to develop strategies to extract scores from these LLM’s that grade machine translations and summaries. We will specifically focus on prompting techniques, therefore, fine-tuning of the LLM’s is not allowed. Based on the submissions, we hope to explore and formalize prompting approaches for open-source LLM-based metrics and, with that, help to improve their correlation to human judgements. As many prompting techniques produce explanations as a side product we hope that this task will also lead to more explainable metrics. Also, we want to evaluate which of the selected open-source models provide the best capabilities as metrics, thus, as a base for fine-tuning. ** Goals ** The shared task has the following goals: Prompting strategies for LLM-based metrics: We want to explore which prompting strategies perform best for LLM-based metrics. E.g., few-shot prompting [*], where examples of other solutions are given in a prompt, chain-of-thought reasoning (CoT) [*], where the model is prompted to provide a multi-step explanation itself, or tree-of-thought prompting [*], where different explanation paths are considered, and the best is chosen. Also, automatic prompt generation might be considered [*]. Numerous other recent works explore further prompting strategies, some of which use multiple evaluation passes. Score aggregation for LLM-based metrics: We also want to explore which strategies best aggregate the model scores from LLM-based metrics. E.g., scores might be extracted as the probability of a paraphrase being created [*], or they could be extracted from LLM output directly [*]. Explainability for LLM-based metrics: We want to analyze whether the metrics that provide the best explanations (for example with CoT) will achieve the highest correlation to human judgements. We assume that this is the case, due to the human judgements being based on fine-grained evaluations themselves (e.g. MQM for machine translation) ** Task Description ** The task will consist of building a reference-free metric for machine translation and/or summarization that predicts sentence-level quality scores constructed from fine-grained scores or error labels. Reference-free means that the metric rates the provided machine translation solely based on the provided source sentence/paragraph, without any additional, human written references. Further, we note that many open-source LLMs have mostly been trained on English data, adding further challenges to the reference-free setup. To summarize, the task will be structured as follows: - We provide a list of allowed LLMs from Huggingface - Participants should use prompting to use these LLMs as metrics for MT and summarization - Fine-tuning of the selected model(s) is not allowed - We will release baselines, which participants might build upon - We will provide a CodaLab dashboard to compare participants' solutions to others We plan to release a CodaLab submission environment together with baselines and dev set evaluation code successively until August 7. We will allow specific models from Huggingface, please refer to the webpage for more details: https://eval4nlp.github.io/2023/shared-task.html Best wishes, The Eval4NLP organizers [*] References are listed on the shared task webpage: https://eval4nlp.github.io/2023/shared-task.html

1 0

Third call for papers DHASA Conference 2023 Extended deadlines
by Menno Van Zaanen 07 Aug '23

07 Aug '23

Third call for papers DHASA Conference 2023 Extended deadlines https://dh2023.digitalhumanities.org.za/ Note extended deadlines Theme: "Digital Humanities for Inclusion" The Digital Humanities Association of Southern Africa (DHASA) is pleased to announce its fourth conference, focusing on the theme "Digital Humanities for Inclusion." In a region where the field of Digital Humanities is still relatively underdeveloped, this conference aims to address this gap and foster growth and collaboration in the field. The conference offers an opportunity for researchers interested in showcasing their work in the broad field of Digital Humanities to come together. By doing so, the conference provides a comprehensive overview of the current state-of-the-art in Digital Humanities, particularly within the Southern Africa region. As such, we welcome submissions related to Digital Humanities research conducted by individuals from Southern Africa or research focused on the geographical area of Southern Africa. Furthermore, the conference serves as a platform for information sharing and networking among researchers passionate about Digital Humanities. By bringing together experts working on Digital Humanities in Southern Africa or with a focus on Southern Africa, we aim to promote collaboration and facilitate further research in this dynamic field. In addition to the main conference, affiliated workshops and tutorials will be organized, providing researchers with valuable insights into novel technologies and tools. These supplementary events are designed for researchers interested in specific aspects of Digital Humanities or seeking practical information to enter or advance their knowledge in the field. The DHASA conference welcomes interdisciplinary contributions from researchers in various domains of Digital Humanities, including, but not limited to, language, literature, visual art, performance and theatre studies, media studies, music, history, sociology, psychology, language technologies, library studies, philosophy, methodologies, software and computation, and more. Our goal is to cultivate an inclusive scientific community of practice within Digital Humanities. Suggested topics include the following: * Digital archives and the preservation of marginalized voices; * Intersectionality and the digital humanities: exploring the intersections of race, gender, sexuality, and class in digital research and activism; * Activism and social change through digital media: how digital humanities tools and methodologies can be used to promote inclusion; * Engaging marginalized communities in the creation and use of digital tools and resources; * Exploring the role of digital humanities in decolonizing knowledge and promoting indigenous perspectives; * The ethics of data collection and analysis in digital humanities research related; * The role of digital humanities in promoting inclusive and equitable pedagogy; * Digital humanities and inclusion in the context of global perspectives and international collaborations; * Critical approaches to digital humanities and inclusion: examining the limitations and possibilities of digital tools and methodologies in promoting inclusion; and * Collaborative digital humanities projects with non-profit organizations, community groups, and cultural institutions; * Any other digital humanities-related topic that serves the Southern African community. Submission Guidelines The DHASA conference 2023 asks for three types of submissions: * Long papers: Authors may submit long papers consisting of a maximum of 8 content pages and unlimited pages for references and appendix. The final versions of accepted long papers will be granted an additional page (up to 9 pages) to incorporate reviewers' comments. * Short papers: Authors may submit short papers with a maximum of 5 content pages and unlimited pages for references and appendix. The final versions of accepted short papers will be allowed an extra page (up to 6 pages) to accommodate reviewers' comments. Short papers accepted for the conference will be presented as posters. * Abstracts: Authors can submit abstracts of 250-300 words. Note that before submitting your contribution, you are required to submit an abstract before the abstract submission deadline. This holds for *all* submissions. The actual submission will need to be submitted before the submission deadline. More information on the submission process can be found on the submission page: https://dh2023.digitalhumanities.org.za/submission/ We particularly encourage student submissions where the first author is a student. All accepted long and short paper submissions that are presented at the conference will be published in the Journal of Digital Humanities Association of Southern Africa, see https://upjournals.up.ac.za/index.php/dhasa. In addition, the abstracts of the full papers and the lightning talks will be published in a book of abstracts before the conference. Important dates Abstract submission deadline: *22 August 2023* Submission deadline: *29 August 2023* Date of notification: 30 September 2023 Camera-ready copy deadline: 6 November 2023 Conference: 27 November 2023 - 1 December 2023 Conference format: Face-to-face Conference venue: Nelson Mandela University, Eastern Cape South Africa NOTE: Non-presenting delegates have the option to attend online. Co-located events Several co-located events are currently being prepared, including workshops and tutorials. These will be updated on the conference website. Organizing Committee * Johannes Sibeko, Nelson Mandela University * Aby Louw, Council for Scientific and Industrial Research * Alan Murdoch, Nelson Mandela University * Amanda du Preez, University of Pretoria * Andiswa Bukula, South African Centre for Digital Language Resources * Andiswa Mvanyashe, Nelson Mandela University * Avashna Govender, Council for Scientific and Industrial Research * Gabby Dlamini, Nelson Mandela University * Ilana Wilken, Council for Scientific and Industrial Research * Jonathan van der Walt, Nelson Mandela University * Laurette Marais, Council for Scientific and Industrial Research * Mukhtar Raban, Nelson Mandela University * Nomfundo Khumalo, Nelson Mandela University * Menno Van Zaanen, South African Centre for Digital Language Resources -- Prof Menno van Zaanen menno.vanzaanen(a)nwu.ac.za Professor in Digital Humanities South African Centre for Digital Language Resources https://www.sadilar.org ________________________________ NWU PRIVACY STATEMENT: http://www.nwu.ac.za/it/gov-man/disclaimer.html DISCLAIMER: This e-mail message and attachments thereto are intended solely for the recipient(s) and may contain confidential and privileged information. Any unauthorised review, use, disclosure, or distribution is prohibited. If you have received the e-mail by mistake, please contact the sender or reply e-mail and delete the e-mail and its attachments (where appropriate) from your system. ________________________________

1 0

Call for participation: SIGDIAL-INLG 2023, Sep 11-15, Prague Czechia
by Ondrej Dusek 07 Aug '23

07 Aug '23

*** Call for Participation *** SIGDIAL & INLG 2023 Conferences September 11-15, 2023 Prague, Czechia & online https://sigdialinlg2023.github.io/ Early Registration Deadline: **August 10** Late Registration Deadline: September 15 Non-presenters Free Registration: August 12 - September 15 Workshops: September 11-12 Main Conferences: September 13-15 The 24th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGdial 2023) and the 16th International Natural Language Generation Conference (INLG 2023) will be held jointly this year in Prague, Czechia. The event will be hybrid, but in-person participation is strongly encouraged! Virtual attendance will be free for non-presenters. The organizers of SIGDIAL & INLG 2023 invite all researchers and practitioners, SIGDial & SIGGEN members, and SIGDIAL & INLG 2023 industry partners and sponsors to join the conference. The registration is now open, with early rates available until August 10 (see https://sigdialinlg2023.github.io/registration.html). A limited number of hotel rooms at the conference venue is available for booking at special rates until August 10. **SIGDIAL** provides a regular forum for the presentation of cutting-edge research in discourse and dialogue to both academic and industry researchers. Continuing a series of 23 successful previous meetings, this conference spans the research interest areas of discourse and dialogue. The conference is sponsored by the SIGdial organization, which serves as the Special Interest Group on discourse and dialogue for both ACL and ISCA. **INLG** is a yearly venue for presentations related to all aspects of Natural Language Generation (NLG), including data-to-text, concept-to-text, text-to-text and vision-to-text approaches. The event is organized under the auspices of SIGGEN, the Special Interest Group on Natural Language Generation of ACL. The joint conference on Sep 13-15 will feature 4 keynote speeches by Barbara Di Eugenio, Emmanuel Dupoux, Elena Simperl and Ryan Lowe, as well as a number of regular paper presentations and system demonstrations. - Keynotes: https://sigdialinlg2023.github.io/speakers.html - SIGDIAL accepted papers: https://2023.sigdial.org/accepted-papers/ - INLG accepted papers: https://inlg2023.github.io/accepted_papers.html The event includes several workshops on Sep 11-12 (see https://sigdialinlg2023.github.io/workshops.html): - YRRSDS: 19th Young Researchers' Roundtable on Spoken Dialogue Systems - The 1st Workshop on Counter Speech for Online Abuse - DSTC11: The 11th Dialog System Technology Challenge - PracticalD2T: 1st Workshop on Practical LLM-assisted Data-to-Text Generation - Taming Large Language Models: Controllability in the era of Interactive Assistants - Workshop on Multimodal, Multilingual Natural Language Generation and Multilingual WebNLG Challenge - Connecting multiple disciplines to AI techniques in interaction-centric autism research and diagnosis - Designing divergent agent tasks for SDS data collection We thank you for your support and look forward to welcoming you at the conference! Best regards, SIGDIAL & INLG 2023 Organizers (Apologies for cross-posting)

1 0

LREC-COLING 2024 website is alive!
by sara goggi 07 Aug '23

07 Aug '23

Dear Corpora members, this is to announce that the LREC COLING 2024 website is now available at: https://lrec-coling-2024.org/ On the website you will find the 2nd Call for Papers for the main conference, the Workshops CfP and the Tutorials CfP, the Author's kit plus other information about Torino. All the best, LREC-COLING 2024 Organizers

1 0

Fwd: looking for summarization datasets in italian and spanish langs
by Saeed Farzi 06 Aug '23

06 Aug '23

Hi guys, I am going to implement a summarization system in the medical domain in Italian and Spanish. So I am looking for free summarization datasets both in the public and medical domains in both languages. Any help would be appreciated. sincerely Ciao -- *Dr. Saeed Farzi,* Faculty of Computer Engineering, K. N. Toosi University of Technology, Tehran, Iran. Phone: +98-21-8462450-401 Fax: +98-21-88462066 P.O. Box: 16315-1355, Web: http://wp.kntu.ac.ir/saeedfarzi/ Lab: https://www.trlab.ir/ -- *Dr. Saeed Farzi,* Faculty of Computer Engineering, K. N. Toosi University of Technology, Tehran, Iran. Phone: +98-21-8462450-401 Fax: +98-21-88462066 P.O. Box: 16315-1355, Web: http://wp.kntu.ac.ir/saeedfarzi/ Lab: https://www.trlab.ir/

2 1

Third CFP (ArabicNLP 2023): The 1st Arabic Natural Language Processing Conference
by Amr Keleg 06 Aug '23

06 Aug '23

*** Apologies for Cross-Posting *** The First Arabic Natural Language Processing Conference (ArabicNLP 2023) co-located with EMNLP 2023 in Singapore. What's in a name? To mark our move from a workshop to a conference, we changed our acronym from WANLP to ArabicNLP. Conference URL: <https://wanlp2023.sigarab.org/> https://arabicnlp2023.sigarab.org/ Submission URL: https://openreview.net/group?id=SIGARAB.org/ArabicNLP/2023/Conference ArabicNLP 2023 invites the submission of original long, short, or demo papers in the area of Arabic Natural Language Processing. ArabicNLP 2023 builds on seven previous workshop editions, which have been extremely successful, drawing in a large active participation in various capacities. This conference is timely given the continued rise in research projects focusing on Arabic NLP. ArabicNLP 2023 will also feature shared tasks, allowing participants to work on specific NLP challenges related to Arabic language processing. The conference is organized by the Special Interest Group on Arabic NLP (SIGARAB), an Association for Computational Linguistics Special Interest Group on Arabic Natural Language Processing. Important Dates - May 7, 2023: submission of shared tasks proposals - May 14, 2023: notification of acceptance of shared tasks - September 5, 2023: conference papers due date - October 12, 2023: notification of acceptance - October 20, 2023: camera-ready papers due - December 7, 2023: conference day All deadlines are 11:59 pm UTC -12h <https://www.timeanddate.com/time/zone/timezone/utc-12> (“Anywhere on Earth”). We accept long (up to 8 pages), short (up to 4 pages), and demo paper (up to 4 pages) submissions. Long and short papers will be presented orally or as posters as determined by the program committee. Submissions are invited on topics that include, but are not limited to, the following: - Enabling core technologies: language models and large language models, morphological analysis, disambiguation, tokenization, POS tagging, named entity detection, chunking, parsing, semantic role labeling, sentiment analysis, Arabic dialect modeling, etc. - Applications: dialog modeling, machine translation, speech recognition, speech synthesis, optical character recognition, pedagogy, assistive technologies, social media, etc. - Resources: dictionaries, annotated data, corpora, etc. Submissions may include work in progress as well as finished work. Submissions must have a clear focus on specific issues pertaining to the Arabic language whether it is standard Arabic, dialectal, classical, or mixed. Papers on other languages sharing problems faced by Arabic NLP researchers, such as Semitic languages or languages using Arabic script, are welcome provided that they propose techniques or approaches that would be of interest to Arabic NLP, and they explain why this is the case. Additionally, papers on efforts using Arabic resources but targeting other languages are also welcome. Descriptions of commercial systems are welcome, but authors should be willing to discuss the details of their work. If you have any questions, please contact us at: <https://groups.google.com/u/1/>arabicnlp-pc-chairs(a)sigarab.org The ArabicNLP 2023 Publicity Chairs, Amr Keleg and Salam Khalifa

1 0

Re: Any literature about tensors-based corpora NLP research with actual examples (and homework ;-)) you would suggest? ...
by Albretch Mueller 05 Aug '23

05 Aug '23

On 8/3/23, Toms Bergmanis <toms.bergmanis(a)tilde.lv> wrote: ... I, for one, have benefited from Ada's, as well as other member's suggestions and comments as I hope they have somehow benefited from mine. lbrtchx

4 7

1st CfP special issue in « Computational Linguistics »
by Abdellah Fourtassi 05 Aug '23

05 Aug '23

1st Call for Papers: Special Issue of the Computational Linguistics journal on Language Learning, Representation, and Processing in Humans and MachinesGuest Editors Marianna Apidianaki (University of Pennsylvania) Abdellah Fourtassi (Aix Marseille University) Sebastian Padó (University of Stuttgart) *Submission deadline: December, 10* Large language models (LLMs) acquire rich world knowledge from the data they are exposed to during training, in a way that appears to parallel how children learn from the language they hear around them. Indeed, since the introduction of these powerful models, there has been a general feeling among researchers in both NLP and cognitive science that a systematic understanding of how these models work and how they use the knowledge they encode, would shed light on the way humans acquire, represent, and process this same knowledge (and vice versa). Yet, despite the similarities, there are important differences between machines and humans that have prevented a direct translation of insights from the analysis of LLMs to a deeper understanding of human learning. Chief among these differences is that the size of data required to train LLMs far exceeds -- by several orders of magnitude -- the data children need to acquire sophisticated conceptual structures and meanings. Besides, the engineering-driven architectures of LLMs do not appear to have obvious equivalents in children's cognitive apparatus, at least as studied by standard methods in experimental psychology. Finally, children acquire world knowledge not only via exposure to language but also via sensory experience and social interaction. This edited volume aims to create a forum of exchange and debate between linguists, cognitive scientists and experts in deep learning, NLP and computational linguistics, on the broad topic of learning in humans and machines. Experts from these communities can contribute with empirical and theoretical papers that advance our understanding of this question. Submissions might address the acquisition of different types of linguistic and world knowledge. Additionally, we invite contributions that characterize and address challenges related to the mismatch between humans and LLMs in terms of the size and nature of input data, and the involved learning and processing mechanisms. Topics include, but are not limited to: - Grounded learning: comparison of unimodal (e.g., text) vs multimodal (e.g., images and video) learning. - Social learning: comparison of input-driven mechanisms vs. interaction-based learning. - Exploration of different knowledge types (e.g., procedural / declarative); knowledge integration and inference in LLMs. - Methods to characterize and quantify human-like language learning or processing in LLMs. - Interpretability/probing methods addressing the linguistic and world knowledge encoded in LLM representations. - Knowledge enrichment methods aimed at improving the quality and quantity of the knowledge encoded in LLMs. - Semantic representation and processing in humans and machines in terms of, e.g., abstractions made, structure of the lexicon, property inheritance and generalization, geometrical approaches to meaning representation, mental associations, and meaning retrieval. - Bilingualism in humans and machines; second language acquisition in children and adults; construction of multi-lingual spaces and cross-lingual correspondences. - Exploration of language models that incorporate cognitively plausible mechanisms and reasonably-sized training data. - Use of techniques from other disciplines (e.g., neuroscience or computer vision) for analyzing and evaluating LLMs. - Open-source tools for analysis, visualization, or explanation. Submission Instructions Papers should be formatted according to the Computational Linguistics style guidelines: https://cljournal.org/ We accept both long and short papers. Long papers are between 25 and 40 journal pages in length; short papers are between 15 and 25 pages in length. Papers for this special issue will be submitted through the CL electronic submission system, just like regular papers: https://cljournal.org/submissions.html Authors of special issue papers will need to select ‟Special Issue on LLRP‟ under the Journal Section heading in the CL submission system. Please note that papers submitted to a special issue undergo the same reviewing process as regular papers. Timeline Deadline for submissions : December, 10 2023 Notification after 1st round of reviewing : February, 10 2024 Revised versions of the papers : April, 30 2024 Final decisions : June, 10 2024 Final version of the papers : July, 1 2024Guest Editors Marianna Apidianaki marapi(a)seas.upenn.edu Abdellah Fourtassi abdellah.fourtassi(a)gmail.com Sebastian Padó pado(a)ims.uni-stuttgart.de *Computational Linguistics* is the longest-running flagship journal of the Association for Computational Linguistics. The journal has a high impact factor: 9.3 in 2022 and 7.778 in 2021. Average time to first decision of regular papers and full survey papers (excluding desk rejects) is 34 days for the period January to May 2023, and 47 days for the period January to December 2022. -- This email was sent from my smartphone. Forgive the brevity, the typos, and the lack of nuance.

1 0

[Last CFP] 3rd Workshop for NLP Open Source Software at EMNLP 2023
by liling tan 04 Aug '23

04 Aug '23

(apologies for cross-posting) *-----Workshop for NLP Open Source Software (NLP-OSS)* 06 Dec 2023, Co-located with EMNLP 2023 https://nlposs.github.io/ Deadline for Long and Short Paper submission: *09 August, 2023 (23:59, GMT-11)* ----- You have tried to use the latest, bestest, fastest LLM models and bore grievances but found the solution after hours of coffee and computer staring. Share that at NLP-OSS and suggest how open source could change for the better (e.g. best practices, documentation, API design etc.) You came across an awesome SOTA system on NLP task X that no LLM has beaten its F1 score. But now the code is stale and it takes a dinosaur to understand the code. Share your experience at NLP-OSS and propose how to "replicate" these forgotten systems. You see this shiny GPT from a blog post, tried it to reproduce similar results on a different task and it just doesn't work on your dataset. You did some magic to the code and now it works. Show us how you did it! Though they're small tweaks, well-motivated and empirically tested are valid submissions to NLP-OSS. You have tried 101 NLP tools and there's none that really do what you want. So you wrote your own shiny new package and made it open source. Tell us why your package is better than the existing tools. How did you design the code? Is it going to be a one-time thing? Or would you like to see thousands of people using it? You have heard enough of open-source LLM and pseudo-open-source GPT but not enough about how it can be used for your use-case or your commercial product at scale. So you contacted your legal department and they explained to you about how data, model and code licenses work. Sharing the knowledge with the NLP-OSS community. You have a position/opinion to share about free vs open vs closed source LLMs and have valid arguments, references or survey/data to support your position. We would like to hear more about it. At last, you've found the avenue to air these issues in an academic platform at the NLP-OSS workshop!!! Sharing your experiences, suggestions and analysis from/of NLP-OSS ---- P/S: 2nd Call for Paper *Workshop for NLP Open Source Software (NLP-OSS)* 06 Dec 2023, Co-located with EMNLP 2023 https://nlposs.github.io/ Deadline for Long and Short Paper submission: 09 August, 2023 (23:59, GMT-11) ------------------------------ The Third Workshop for NLP Open Source Software (NLP-OSS) will be co-located with EMNLP 2023 on 06 Dec 2023. Focusing more on the social and engineering aspect of NLP software and less on scientific novelty or state-of-art models, the Workshop for NLP-OSS is an academic forum to advance open source developments for NLP research, teaching and application. NLP-OSS also provides an academic workshop to announce new software/features, promote the collaborative culture and best practices that go beyond the conferences. 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 - Backwards 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 Submission should be formatted according to the EMNLP 2023 templates <https://2023.emnlp.org/call-for-papers> and submitted to OpenReview <https://openreview.net/group?id=EMNLP/2023/Workshop/NLP-OSS> ORGANIZERS Geeticka Chauhan, Massachusetts Institute of Technology Dmitrijs Milajevs, Grayscale AI Elijah Rippeth, University of Maryland Jeremy Gwinnup, Air Force Research Laboratory Liling Tan, Amazon

1 0

2026

2025

2024

2023

2022

Corpora