Mining Feature-based Opinion Expressions by Mutual Information Approach

Su, Qi; Zhu, Ya Jun; Swen, Bin; Yu, Shiwen

doi:10.1142/s0219427907001639

Cited by 3 publications

(1 citation statement)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This method of determining the polarity of a review is simple and fast, but may become too general. Another way to detect the polarity of a review is to examine the textual content of the review (Ku and Chen 2007;Su et al 2007). In this manner, a keyword dictionary generally needs to be established before we can analyze the polarity of reviews using text mining techniques.…”

Section: Problem Descriptionmentioning

confidence: 99%

A sales forecasting model for consumer products based on the influence of online word-of-mouth

Chern

Wei

Fangyi

et al. 2014

Inf Syst E-Bus Manage

View full text Add to dashboard Cite

Sales forecasting is one of the most critical steps of business process. Since the forecasting accuracy of traditional techniques is generally unacceptable for products with irregular or non-seasonal sales trends, it is necessary to construct a new forecasting method. Past research shows that there is a strong relationship between online word-of-mouth and product sales, but that the extent of the impact of word-of-mouth varies with product category. This study aims to provide an understanding of how electronic word-of-mouth affects product sales by analyzing online review properties, reviewer characteristics and review influences. This new electronic word-of-mouth perspective contributes to sales forecasting research in two ways. First, a novel classification model involving polarity mining, intensity mining and influence analysis is proposed with a framework to elucidate the difference between review categories. Second, the influence of online reviews (i.e., electronic word-of-mouth) is estimated and then used to construct a sales forecasting model. The proposed online word-of-mouth-based sales forecasting method is evaluated by using real data from a well-known cosmetic retail chain in Taiwan. The experimental results demonstrate that the proposed method is especially suitable for products with abundant online reviews and outperforms traditional time series forecasting models for most consumer products examined.

show abstract

Section: Problem Descriptionmentioning

confidence: 99%