Search

Extracting the user reviews in websites such as forums, blogs, newspapers, commerce, trips, etc. is crucial for text processing applications (e.g. sentiment analysis, trend detection/monitoring and recommendation systems) which are needed to deal with structured data. Traditional algorithms have three processes consisting of Document Object Model (DOM) tree creation, extraction of features obtained from this tree and machine learning. However, these algorithms increase time complexity of extraction process. This study proposes a novel algorithm that involves two complementary stages. The first stage determines which HTML tags correspond to review layout for a web domain by using the DOM tree as well as its features and decision tree learning. The second stage extracts review layout for web pages in a web domain using the found tags obtained from the first stage. This stage is more time-efficient, being approximately 21 times faster compared to the first stage. Moreover, it achieves a relatively high accuracy of 96.67% in our experiments of review block extraction. [ABSTRACT FROM AUTHOR]

Recently, it has become easy to collect large volumes of customers' voice through the web as text data. Accordingly, the role of text mining has become more and more important for various business situations to obtain useful knowledge easily and in large quantities. However, the technique of visualizing the correspondence relation between customer reviews and evaluation information is insufficient. The purpose of this paper is to propose a new method of visualizing the information from the text data of a customer review using a self-organizing map robust for non-linearity and multi-collinearity. In this proposal, probabilistic latent semantic indexing, which does not require weighting for dimension contraction of a word vector, was used. Furthermore, the method of visualizing the distribution of evaluation information in a self-organizing map is newly proposed. Here, in order to give a suitable value to the nodes without an evaluation value and the dead nodes, an interpolation formula for the evaluation value was newly defined. To confirm the validity of our proposal, we visualized the customer review data in a major electronic commerce site using the proposed method. [ABSTRACT FROM AUTHOR]