使用回归分析策略以文档满足用户的信息需求程度作为回归分析的目标值,利用回归支持向量机构建了信息检索模型.新模型不仅提供了融合不同来源特征的灵活框架,而且由于使用回归支持向量机寻找具有ε不敏感损失的回归函数,因此具有良好的泛化性能.实验表明,新模型性能优于目前主流的基于语言模型的信息检索方法.
The regression method is explored for IR, and the degree is used as regression target value. Support vector machine regression (SVMR)is adopted in the framework because it provides a flexible framework to incorporate arbitrary features. SVMR is used to find a regression function with ε insensitive loss, which allows good generalization. Results show that the new model outperforms the state-of-the-art language modeling approaches.