arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.20336cs.CL
收藏

用于鲁棒文档分类的自然语言处理模型

Natural Language Processing Models for Robust Document Categorization

Faculty of Electrical Engineering, Warsaw University of Technology(华沙技术大学电气工程学院);Warsaw University of Technology(华沙技术大学)

Radoslaw Roszczyk, Pawel Tecza, Maciej Stodolski, Krzysztof Siwek

AI总结:

本文提出BiLSTM模型在不平衡文档分类中实现最佳平衡,兼顾准确率与效率。

摘要:

本文介绍了几种机器学习方法应用于自动文本分类的评估,同时设计了一个演示系统用于不平衡文档分类和分发。研究重点在于平衡分类准确率与计算效率,这是在将AI整合到现实世界自动化流水线中的关键考虑因素。三种复杂度不同的模型被检验:一个朴素贝叶斯分类器,一个双向LSTM网络,以及一个微调的基于BERT的Transformer模型。实验结果显示性能存在显著差异。BERT实现了最高准确率,持续超过99%,但需要显著更长的训练时间和更大的计算资源。BiLSTM模型提供了良好的折中方案,达到约98.56%的准确率,同时保持中等的训练成本,并提供强大的上下文理解能力。朴素贝叶斯证明了训练速度最快,大约在毫秒级别,但交付的准确率最低,平均约为94.5%。类别不平衡影响了所有方法,特别是在少数类别的识别上。一个完整的演示系统被实现以验证实际适用性,使技术请求的自动路由成为可能,其吞吐量无法通过手动处理实现。研究结论指出BiLSTM为所研究的场景提供了最平衡的解决方案,同时概述了未来改进和进一步探索Transformer架构的机会。

英文摘要:

This article presents an evaluation of several machine learning methods applied to automated text classification, alongside the design of a demonstrative system for unbalanced document categorization and distribution. The study focuses on balancing classification accuracy with computational efficiency, a key consideration when integrating AI into real world automation pipelines. Three models of varying complexity were examined: a Naive Bayes classifier, a bidirectional LSTM network, and a fine tuned transformer based BERT model. The experiments reveal substantial differences in performance. BERT achieved the highest accuracy, consistently exceeding 99\%, but required significantly longer training times and greater computational resources. The BiLSTM model provided a strong compromise, reaching approximately 98.56\% accuracy while maintaining moderate training costs and offering robust contextual understanding. Naive Bayes proved to be the fastest to train, on the order of milliseconds, yet delivered the lowest accuracy, averaging around 94.5\%. Class imbalance influenced all methods, particularly in the recognition of minority categories. A fully functional demonstrative system was implemented to validate practical applicability, enabling automated routing of technical requests with throughput unattainable through manual processing. The study concludes that BiLSTM offers the most balanced solution for the examined scenario, while also outlining opportunities for future improvements and further exploration of transformer architectures.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑