paper-with-me

Papers

Incremental user embedding modeling for personalized text classification

2022-02-13 · Ruixue Lian, Che-Wei Huang, Yuqing Tang, Qilong Gu, Chengyuan Ma, Chenlei Guo

Individual user profiles and interaction histories play a significant role in providing customized experiences in real-world applications such as chatbots, social media, retail, and education. Adaptive user representation learning by utilizing user personalized information has become increasingly challenging due to ever-growing history data. In this work, we propose an incremental user embedding modeling approach, in which embeddings of user's recent interaction histories are dynamically integrated into the accumulated history vectors via a transformer encoder. This modeling paradigm allows us to create generalized user representations in a consecutive manner and also alleviate the challenges of data management. We demonstrate the effectiveness of this approach by applying it to a personalized multi-class classification task based on the Reddit dataset, and achieve 9% and 30% relative improvement on prediction accuracy over a baseline system for two experiment settings through appropriate comment history encoding and task modeling.

📄 PDF Abstract BibTeX arXiv:2202.06369

Code (0)

등록된 구현이 없습니다.

Tasks

ClassificationManagementMulti-class ClassificationRepresentation Learningtext-classificationText Classification

Similar Papers 제목 키워드 기반

User Embedding Model for Personalized Language Prompting

2024-01-10 · Sumanth Doddapaneni, Krishna Sayana, Ambarish Jash, Sukhdeep Sodhi 외

Modeling long histories plays a pivotal role in enhancing recommendation systems, allowing to capture user's evolving preferences, resulting in more precise and personalized recommendations. In this study we tackle the c…

modelRecommendation Systems

HieRec: Hierarchical User Interest Modeling for Personalized News Recommendation

2021-06-08 · ACL 2021 5 · Tao Qi, Fangzhao Wu, Chuhan Wu, Peiru Yang 외

User interest modeling is critical for personalized news recommendation. Existing news recommendation methods usually learn a single user embedding for each user from their previous behaviors to represent their overall i…

News Recommendation

LLMs + Persona-Plug = Personalized LLMs

2024-09-18 · Jiongnan Liu, Yutao Zhu, Shuting Wang, Xiaochi Wei 외

Personalization plays a critical role in numerous language tasks and applications, since users with the same requirements may prefer diverse outputs based on their individual interests. This has led to the development of…

Language Modelling

UQABench: Evaluating User Embedding for Prompting LLMs in Personalized Question Answering

2025-02-26 · Langming Liu, Shilei Liu, Yujin Yuan, Yizhen Zhang 외

Large language models (LLMs) achieve remarkable success in natural language processing (NLP). In practical scenarios like recommendations, as users increasingly seek personalized experiences, it becomes crucial to incorp…

Question Answering

Compositional Demographic Word Embeddings

2020-10-06 · EMNLP 2020 11 · Charles Welch, Jonathan K. Kummerfeld, Verónica Pérez-Rosas, Rada Mihalcea

Word embeddings are usually derived from corpora containing text from many individuals, thus leading to general purpose representations rather than individually personalized representations. While personalized embeddings…

Language ModelingLanguage ModellingWord Embeddings