开源推理模型DeepSeek-R1在情感分析中实现高精度与可解释性
Explainable Sentiment Analysis with DeepSeek-R1: Performance, Efficiency, and Few-Shot Learning
- 对比GPT系列，用少样本训练快速达到高准确率
- 5样本下5分类任务F1达91.39%，效率比GPT-4o快8倍
- 生成清晰推理步骤，适合需要可解释性的场景
大型语言模型（LLMs）已改变情感分析格局，但如何兼顾准确性、效率与可解释性仍是关键挑战。本研究首次全面评估开源推理模型DeepSeek-R1，对比OpenAI的GPT-4o和GPT-4o-mini。测试了完整的671B模型及其蒸馏变体，系统记录少样本学习曲线。实验显示，DeepSeek-R1在5分类情感任务上取得91.39%的F1分数，在二分类任务中达99.31%准确率，仅需5个样本即达成，少样本效率比GPT-4o高出八倍。架构相关的蒸馏效果显现：基于Qwen2.5的32B模型优于基于Llama的70B版本，高出6.69个百分点。尽管推理过程降低吞吐量，但其透明的逐步推理解释提供了更强的可解释性，确立其为强大且可解释的开源替代方案。
原文摘要 · Abstract (English)
Large language models (LLMs) have transformed sentiment analysis, yet balancing accuracy, efficiency, and explainability remains a critical challenge. This study presents the first comprehensive evaluation of DeepSeek-R1--an open-source reasoning model--against OpenAI's GPT-4o and GPT-4o-mini. We test the full 671B model and its distilled variants, systematically documenting few-shot learning curves. Our experiments show DeepSeek-R1 achieves a 91.39\% F1 score on 5-class sentiment and 99.31\% accuracy on binary tasks with just 5 shots, an eightfold improvement in few-shot efficiency over GPT-4o. Architecture-specific distillation effects emerge, where a 32B Qwen2.5-based model outperforms the 70B Llama-based variant by 6.69 percentage points. While its reasoning process reduces throughput, DeepSeek-R1 offers superior explainability via transparent, step-by-step traces, establishing it as a powerful, interpretable open-source alternative.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。