Research map / Label-efficient and robust learning / Recognition and long-tail
Personalized Microblog Sentiment Classification via Multi-Task Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| combine global and user-specific classifiers in one personalized model (ablation) | PMSC · Twitter and Weibo | accuracy | Helped |
| add social relation regularization (ablation) | PMSC (global+individual) · Twitter and Weibo | accuracy | Helped |
| use PMSC-Log instead of LFM | personalized sentiment classifier · Twitter 100% train | accuracy: 0.8207 -> 0.8433 | Helped |
| use PMSC-Log instead of RMTL | multi-task learning · Twitter 100% train | accuracy: 0.8180 -> 0.8433 | Helped |
| use PMSC-Log instead of LR-global | logistic regression baseline · Weibo 100% train | accuracy: 0.7906 -> 0.8250 | Helped |
| use PMSC-Log instead of LR-individual | logistic regression baseline · Twitter 100% train | accuracy: 0.8042 -> 0.8433 | Helped |
| use PMSC-SVM instead of PMSC-Log | PMSC · Weibo 50% train | accuracy: 0.8110 -> 0.8140 | Mixed |
| set social regularization weight alpha moderately in [0.1, 1] rather than very small or large (ablation) | PMSC · Twitter | accuracy | Mixed |
| use 4 parallel nodes instead of 1 in distributed ADMM | PMSC · Twitter | running time | Helped |
| use squared or log loss (FISTA-based) instead of hinge loss | PMSC · Twitter | running time | Helped |
| use 100% instead of 50% of training data | PMSC-Log · Twitter | accuracy: 0.8350 -> 0.8433 | Helped |
Closest papers by meaning
- A Neural Group-wise Sentiment Analysis Model with Data Sparsity Awareness (AAAI 2021 · similarity 0.83)
- FINE: Factorized Multimodal Sentiment Analysis via Mutual INformation Estimation (AAAI 2026 · similarity 0.83)
- Semisupervised Autoencoder for Sentiment Analysis (AAAI 2016 · similarity 0.82)
- Microsummarization of Online Reviews: An Experimental Study (AAAI 2016 · similarity 0.82)
- Learning Sentiment-Specific Word Embedding via Global Sentiment Representation (AAAI 2018 · similarity 0.82)
- Practical Sentiment Analysis for Education: The Power of Student Crowdsourcing (AAAI 2024 · similarity 0.80)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.