FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4311431702> ?p ?o ?g. }

• W4311431702 abstract "Natural language processing (NLP) sees rich mobile applications. To support various language understanding tasks, a foundation NLP model is often fine-tuned in a federated, privacy-preserving setting (FL). This process currently relies on at least hundreds of thousands of labeled training samples from mobile clients; yet mobile users often lack willingness or knowledge to label their data. Such an inadequacy of data labels is known as a few-shot scenario; it becomes the key blocker for mobile NLP applications. For the first time, this work investigates federated NLP in the few-shot scenario (FedFSL). By retrofitting algorithmic advances of pseudo labeling and prompt learning, we first establish a training pipeline that delivers competitive accuracy when only 0.05% (fewer than 100) of the training data is labeled and the remaining is unlabeled. To instantiate the workflow, we further present a system FeS, addressing the high execution cost with novel designs. (1) Curriculum pacing, which injects pseudo labels to the training workflow at a rate commensurate to the learning progress; (2) Representational diversity, a mechanism for selecting the most learnable data, only for which pseudo labels will be generated; (3) Co-planning of a model's training depth and layer capacity. Together, these designs reduce the training delay, client energy, and network traffic by up to 46.0$times$, 41.2$times$ and 3000.0$times$, respectively. Through algorithm/system co-design, FFNLP demonstrates that FL can apply to challenging settings where most training samples are unlabeled." @default.
• W4311431702 created "2022-12-26" @default.
• W4311431702 creator A5001045957 @default.
• W4311431702 creator A5025585492 @default.
• W4311431702 creator A5054814598 @default.
• W4311431702 creator A5070424477 @default.
• W4311431702 creator A5080027119 @default.
• W4311431702 date "2023-10-02" @default.
• W4311431702 modified "2023-10-01" @default.
• W4311431702 title "Federated Few-Shot Learning for Mobile NLP" @default.
• W4311431702 cites W2510725918 @default.
• W4311431702 cites W2558881939 @default.
• W4311431702 cites W2604230453 @default.
• W4311431702 cites W2612925715 @default.
• W4311431702 cites W2743151379 @default.
• W4311431702 cites W2767079719 @default.
• W4311431702 cites W2779457220 @default.
• W4311431702 cites W2917049430 @default.
• W4311431702 cites W2963460174 @default.
• W4311431702 cites W2964105864 @default.
• W4311431702 cites W2964236337 @default.
• W4311431702 cites W2970641574 @default.
• W4311431702 cites W3026182744 @default.
• W4311431702 cites W3093755649 @default.
• W4311431702 cites W3099771192 @default.
• W4311431702 cites W3099793224 @default.
• W4311431702 cites W3105122387 @default.
• W4311431702 cites W3108032709 @default.
• W4311431702 cites W3154335119 @default.
• W4311431702 cites W3154608090 @default.
• W4311431702 cites W3155160971 @default.
• W4311431702 cites W3161051256 @default.
• W4311431702 cites W3173617765 @default.
• W4311431702 cites W3185341429 @default.
• W4311431702 cites W3209727316 @default.
• W4311431702 cites W3210103168 @default.
• W4311431702 cites W3212066318 @default.
• W4311431702 cites W3216866458 @default.
• W4311431702 cites W4206648492 @default.
• W4311431702 cites W4210968720 @default.
• W4311431702 cites W4224317517 @default.
• W4311431702 cites W4282974189 @default.
• W4311431702 cites W4283032505 @default.
• W4311431702 cites W4288086191 @default.
• W4311431702 cites W4306178406 @default.
• W4311431702 cites W4306178637 @default.
• W4311431702 cites W4310629087 @default.
• W4311431702 cites W4368353189 @default.
• W4311431702 doi "https://doi.org/10.1145/3570361.3613277" @default.
• W4311431702 hasPublicationYear "2023" @default.
• W4311431702 type Work @default.
• W4311431702 citedByCount "0" @default.
• W4311431702 crossrefType "proceedings-article" @default.
• W4311431702 hasAuthorship W4311431702A5001045957 @default.
• W4311431702 hasAuthorship W4311431702A5025585492 @default.
• W4311431702 hasAuthorship W4311431702A5054814598 @default.
• W4311431702 hasAuthorship W4311431702A5070424477 @default.
• W4311431702 hasAuthorship W4311431702A5080027119 @default.
• W4311431702 hasBestOaLocation W43114317022 @default.
• W4311431702 hasConcept C108583219 @default.
• W4311431702 hasConcept C111919701 @default.
• W4311431702 hasConcept C119857082 @default.
• W4311431702 hasConcept C136764020 @default.
• W4311431702 hasConcept C154945302 @default.
• W4311431702 hasConcept C177212765 @default.
• W4311431702 hasConcept C186967261 @default.
• W4311431702 hasConcept C199360897 @default.
• W4311431702 hasConcept C204321447 @default.
• W4311431702 hasConcept C26517878 @default.
• W4311431702 hasConcept C38652104 @default.
• W4311431702 hasConcept C41008148 @default.
• W4311431702 hasConcept C43521106 @default.
• W4311431702 hasConcept C77088390 @default.
• W4311431702 hasConcept C98045186 @default.
• W4311431702 hasConceptScore W4311431702C108583219 @default.
• W4311431702 hasConceptScore W4311431702C111919701 @default.
• W4311431702 hasConceptScore W4311431702C119857082 @default.
• W4311431702 hasConceptScore W4311431702C136764020 @default.
• W4311431702 hasConceptScore W4311431702C154945302 @default.
• W4311431702 hasConceptScore W4311431702C177212765 @default.
• W4311431702 hasConceptScore W4311431702C186967261 @default.
• W4311431702 hasConceptScore W4311431702C199360897 @default.
• W4311431702 hasConceptScore W4311431702C204321447 @default.
• W4311431702 hasConceptScore W4311431702C26517878 @default.
• W4311431702 hasConceptScore W4311431702C38652104 @default.
• W4311431702 hasConceptScore W4311431702C41008148 @default.
• W4311431702 hasConceptScore W4311431702C43521106 @default.
• W4311431702 hasConceptScore W4311431702C77088390 @default.
• W4311431702 hasConceptScore W4311431702C98045186 @default.
• W4311431702 hasFunder F4320334978 @default.
• W4311431702 hasFunder F4320335777 @default.
• W4311431702 hasLocation W43114317021 @default.
• W4311431702 hasLocation W43114317022 @default.
• W4311431702 hasOpenAccess W4311431702 @default.
• W4311431702 hasPrimaryLocation W43114317021 @default.
• W4311431702 hasRelatedWork W2795261237 @default.
• W4311431702 hasRelatedWork W3014300295 @default.
• W4311431702 hasRelatedWork W3164822677 @default.
• W4311431702 hasRelatedWork W4223943233 @default.
• W4311431702 hasRelatedWork W4225161397 @default.

• next