Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Gong, Jiahui, Ding, Jingtao, Meng, Fanjin, Chen, Guilong, Chen, Hong, Zhao, Shen, Lu, Haisheng, Li, Yong
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2408.09815
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929463701274624
author Gong, Jiahui
Ding, Jingtao
Meng, Fanjin
Chen, Guilong
Chen, Hong
Zhao, Shen
Lu, Haisheng
Li, Yong
author_facet Gong, Jiahui
Ding, Jingtao
Meng, Fanjin
Chen, Guilong
Chen, Hong
Zhao, Shen
Lu, Haisheng
Li, Yong
contents Mobile devices, especially smartphones, can support rich functions and have developed into indispensable tools in daily life. With the rise of generative AI services, smartphones can potentially transform into personalized assistants, anticipating user needs and scheduling services accordingly. Predicting user intents on smartphones, and reflecting anticipated activities based on past interactions and context, remains a pivotal step towards this vision. Existing research predominantly focuses on specific domains, neglecting the challenge of modeling diverse event sequences across dynamic contexts. Leveraging pre-trained language models (PLMs) offers a promising avenue, yet adapting PLMs to on-device user intent prediction presents significant challenges. To address these challenges, we propose PITuning, a Population-to-Individual Tuning framework. PITuning enhances common pattern extraction through dynamic event-to-intent transition modeling and addresses long-tailed preferences via adaptive unlearning strategies. Experimental results on real-world datasets demonstrate PITuning's superior intent prediction performance, highlighting its ability to capture long-tailed preferences and its practicality for on-device prediction scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2408_09815
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Population-to-individual Tuning Framework for Adapting Pretrained LM to On-device User Intent Prediction
Gong, Jiahui
Ding, Jingtao
Meng, Fanjin
Chen, Guilong
Chen, Hong
Zhao, Shen
Lu, Haisheng
Li, Yong
Machine Learning
Human-Computer Interaction
Mobile devices, especially smartphones, can support rich functions and have developed into indispensable tools in daily life. With the rise of generative AI services, smartphones can potentially transform into personalized assistants, anticipating user needs and scheduling services accordingly. Predicting user intents on smartphones, and reflecting anticipated activities based on past interactions and context, remains a pivotal step towards this vision. Existing research predominantly focuses on specific domains, neglecting the challenge of modeling diverse event sequences across dynamic contexts. Leveraging pre-trained language models (PLMs) offers a promising avenue, yet adapting PLMs to on-device user intent prediction presents significant challenges. To address these challenges, we propose PITuning, a Population-to-Individual Tuning framework. PITuning enhances common pattern extraction through dynamic event-to-intent transition modeling and addresses long-tailed preferences via adaptive unlearning strategies. Experimental results on real-world datasets demonstrate PITuning's superior intent prediction performance, highlighting its ability to capture long-tailed preferences and its practicality for on-device prediction scenarios.
title A Population-to-individual Tuning Framework for Adapting Pretrained LM to On-device User Intent Prediction
topic Machine Learning
Human-Computer Interaction
url https://arxiv.org/abs/2408.09815