ARTÍCULO
TITULO

A Rule Extraction Study from SVM on Sentiment Analysis

Guido Bologna and Yoichi Hayashi    

Resumen

A natural way to determine the knowledge embedded within connectionist models is to generate symbolic rules. Nevertheless, extracting rules from Multi Layer Perceptrons (MLPs) is NP-hard. With the advent of social networks, techniques applied to Sentiment Analysis show a growing interest, but rule extraction from connectionist models in this context has been rarely performed because of the very high dimensionality of the input space. To fill the gap we present a case study on rule extraction from ensembles of Neural Networks and Support Vector Machines (SVMs), the purpose being the characterization of the complexity of the rules on two particular Sentiment Analysis problems. Our rule extraction method is based on a special Multi Layer Perceptron architecture for which axis-parallel hyperplanes are precisely located. Two datasets representing movie reviews are transformed into Bag-of-Words vectors and learned by ensembles of neural networks and SVMs. Generated rules from ensembles of MLPs are less accurate and less complex than those extracted from SVMs. Moreover, a clear trade-off appears between rules? accuracy, complexity and covering. For instance, if rules are too complex, less complex rules can be re-extracted by sacrificing to some extent their accuracy. Finally, rules can be viewed as feature detectors in which very often only one word must be present and a longer list of words must be absent.

 Artículos similares

       
 
Yunfei Zhang, Zexu Zhang, Jincai Huang, Tingting She, Min Deng, Hongchao Fan, Peng Xu and Xingshen Deng    
With the rapid development of urban traffic, accurate and up-to-date road maps are in crucial demand for daily human life and urban traffic control. Recently, with the emergence of crowdsourced mapping, a surge in academic attention has been paid to gene... ver más

 
Xiaoqian Cheng, Chengming Li, Weibing Du, Jianming Shen and Zhaoxin Dai    

 
Xuehua Han and Juanle Wang    
Web text, using natural language to describe a disaster event, contains a considerable amount of disaster information. Automatic extraction from web text of this disaster information (e.g., time, location, casualties, and disaster losses) is an important... ver más

 
Building design review is the procedure of checking a design against codes and standard provisions to satisfy the accuracy of the design and identify non-compliances before construction begins. The current approaches for conducting the design review proc... ver más
Revista: Buildings

 
Guido Bologna and Yoichi Hayashi    
A natural way to determine the knowledge embedded within connectionist models is to generate symbolic rules. Nevertheless, extracting rules from Multi Layer Perceptrons (MLPs) is NP-hard. With the advent of social networks, techniques applied to Sentimen... ver más