Anzeige
Erschienen in:
2015 | OriginalPaper | Buchkapitel
Feature Level Mining of Online Reviews Based on a Semi-Supervised Learning Model
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
Online reviews are written by customers based on personal usage experience. They not only help manufacturers better understand consumer responses to their products, but also serve as a reliable source of information help other customers make purchase decision. In this paper, we propose a novel semi-supervised learning algorithm to address the feature-level reviews mining problem. The proposed method consists of three phases: (1) build a support function that characterizes the support of a multi-dimensional distribution of a given data set; (2) decompose a whole data space into a small number of separate clustered regions via a dynamical system associated with the constructed support function; (3) assign a class label to each decomposed region using the information of their constituent labeled data and the constructed dynamical system, thereby classifying in-sample unlabeled data as well as unknown out-of-sample data.