FLEXICO: Sustainable Machine Translation via Self-AdaptationFULL
Machine Translation (MT) is the backbone of a plethora of systems and applications that are present in users’ everyday lives. Despite the research efforts and progress in the MT domain, translation remains a challenging task and MT systems struggle when translating rare words, named entities, domain-specific terminology, idiomatic expressions and culturally specific terms. Thus, to meet the translation performance expectations of the users, engineers are tasked with periodically updating (fine-tuning) MT models to guarantee high translation quality. However, with ever-growing machine learning models, fine-tuning operations become increasingly more expensive, raising serious concerns from a sustainability perspective. Furthermore, not all fine-tunings are guaranteed to lead to increased translation quality, thus corresponding to wasted compute resource.
To address this issue and enhance the sustainability of MT systems, we present FLEXICO, a new approach to engineer self-adaptive MT systems, which leverages (i) black-box predictors to estimate the expected benefits of fine-tuning MT models; and (ii) probabilistic model checking techniques to automate the reasoning about when the benefits of fine-tuning outweigh its costs. Our empirical evaluation on two MT models and language-pairs and across up to 9 domains demonstrates the predictive performance of the black-box models that estimate the expected benefits of fine-tuning, as well as their domain-generalizability. Finally, we show that FLEXICO optimizes system utility when compared to naive baselines, decreasing the number of fine-tunings required to achieve high translation quality.
Tue 29 AprDisplayed time zone: Eastern Time (US & Canada) change
14:00 - 15:30 | Session 7: ApplicationsResearch Track / Artifact Track at 204 Chair(s): Liliana Pasquale University College Dublin & Lero | ||
14:00 25mTalk | FLEXICO: Sustainable Machine Translation via Self-AdaptationFULL Research Track Maria Casimiro Instituto Superior Técnico, Universidade de Lisboa & S3D, Carnegie Mellon University, Paolo Romano IST/INESC-ID, José Sousa Unbabel, Amin M Khan INESC-ID. Universidade de Lisboa, David Garlan Carnegie Mellon University | ||
14:25 25mTalk | SPARQ: A QoS-aware Framework for Mitigating Cyber Risk in Self-Protecting IoT SystemsFULLBest Paper Award Research Track Alessandro Palma Università di Roma Sapienza, Houssam Hajj Hassan SAMOVAR, Télécom SudParis, Institut Polytechnique de Paris, Georgios Bouloukakis Télécom SudParis, Institut Polytechnique de Paris | ||
14:50 15mTalk | Adapting Aggregation Rule for Robust Federated Learning under Dynamic AttacksSHORT Research Track Chenyu Hu Southwest University, Mingyue Zhang Southwest University, NIANYU LI ZGC Lab, China, Jialong Li Waseda University, Japan, Zheng Yang Southwest University, Muneeb Ul Hassan Deakin University, Kenji Tei Institute of Science Tokyo | ||
15:05 15mTalk | Adaptive and Interoperable Federated Data Spaces: An Implementation ExperienceARTIFACT Artifact Track Nikolaos Papadakis , Niemat Khoder Télécom SudParis, Institut Polytechnique de Paris, France, Daphne Tuncer Ecole nationale des ponts et chaussees, Institut Polytechnique de Paris, France, Kostas Magoutis University of Crete and FORTH-ICS, Georgios Bouloukakis Télécom SudParis, Institut Polytechnique de Paris | ||
15:20 10mOther | Discussion Session 7 Research Track |