Li Yi-jun

Sentiment classification for Chinese product reviews using an unsupervised Internet-based method

Sentiment classification aims at mining opinions of customers for a certain product by automatically classifying the reviews into positive or negative opinions. With the fast developing of World Wide Web applications, sentiment classification would have huge opportunity to help automatic analysis of customerspsila opinions from the web information. Opinion mining will benefit both consumers and sellers. Up to now, it is still a complicated task with great challenge. Though some pioneer researches explored the approaches for English review classification, few works have been done on sentiment classification for Chinese reviews. In this paper, we focus on a specific domain-cell phone review and propose an Internet-based approach for Chinese product review mining. The experimental results show the effectiveness of the proposed approach in sentiment classification for Chinese product reviews.

Exploitation vs. exploration: Choosing keywords for search-based advertising services

Search-based advertising services provided by search engines enable companies to promote their products to internet users based on their queries is now a major online advertising channel. In most search based advertising services, a company sets a budget, selects keywords, bid for each keyword and pay for each click received. Thus the profit of the company highly lies on the reward of each keyword, i.e. conversion rate. With many keywords available and uncertain conversion rate associated with the ad for each keyword, identifying the most profitable keywords becomes challenge for companies. In this research, we consider the challenge as the exploitation and exploration trade-off problem which advertisers must choose between maximizing its expected reward according to its current knowledge about the keywords and trying to learn more knowledge about other keywords, since this might find keywords with higher reward. We first show the similarity of the keyword problem to the multi-armed bandit problem and use Gittins indices to solve the problem with 1-keyword per stage. We also propose four policies for the problem with multiple keywords per stage and suggest the best policy for different context by simulation.

A Comprehensive Evaluation Method of E-commerce Websites Using GA, CA and AHP

The evaluation of e-commerce Web sites is the main measure of knowing the value, comparing the quality and guiding the construction of the Web sites. With the increasing use of e-commerce Web sites, it is gaining more and more attention, and becoming the focus of study. Of all the evaluation methods, analytic hierarchy process (AHP) is favored by many researchers because of its strict mathematical principles, and considering comprehensive features at the same time. But as we all know, in the process of using AHP we should set up a comparison matrix by comparing pairs of criteria or subcriteria. When there are many Web sites involved, it is very possible that the compare results disagree with each other. However, the two new methods, grey analysis (GA) and concordance analysis (CA) demonstrate their advantages for their demand of comparably few data, simple calculation, and straightaway principles. Furthermore the two methods are easy to achieve programming. This paper presents that in the process of evaluation, we can first apply the two methods to evaluate the e-commerce Web sites. If the results of the two evaluations are the same, we accept the results. Otherwise we use AHP to select the better one of the best results acquired by using the above two methods. By using the combination of the three methods, we acquire a more objective and believable evaluation result. In the fourth part of the paper, we prove that the result of this comprehensive method is the same as only using AHP by applying it to a case. And this method would be much easier when it achieves programming in the future

Dynamic impacts of online reviews and other information sources on sales in panel data environment: Evidence from movie industry

In recent years, with the significantly increasing impact of online reviews (a form of word-of-mouth) on consumer purchase decision, clarifying how this kind of information source influences consumer behaviors(usually represented by sales) have been exposed to more attentions in business and academic research fields. Differing from most existing studies which statically analyze relations between online reviews and sales with cross-section data, this paper, based on the movie industry context, employs panel data to dynamically explore the differences of the relationship between online reviews (mainly Volume and Valence) and movie box office revenues at moviespsila different life stages. Meanwhile, this paper further comparatively analyzes the variances of impacts of three information sources (including online reviews, critic reviews and movie production budgets) on movie box offices. Results show that, in panel data environment, the impacts of Volume and Valence of online reviews on movie box office revenues are varying and even mixed for moviespsila different life stages. The effects of Volume and Valence of online reviews on box offices are both insignificant in the movie opening week, while the other two information sources critic reviews and movie production budgets show positive influences on box offices, especially critic reviews exerting stronger impact. From the second week, the positive impact of online reviews Volume becomes significant, but the power of this effect declines in the following weeks. Our results also show that online reviews Valence positively impacts box offices in the third and fourth week, furthermore, the impact power exceeds the other two information sources.

Research on VRP of Optimization Based on Improved Two Phase Algorithm under Electronic Commerce

With the rapid development of electronic commerce, as one of bottlenecks of electronic commerce development, the logistic distribution arouses wide attention. Vehicle routing problem (VRP) plays key role in logistic distribution system. The logistic distribution under electronic commerce environment has the character of dispersive customer positions, large order forms, little batches and many repeated routes. The traditional optimizing route methods meet with diversified problems in different extents and are difficult to play their roles. Therefore, according to the particularity of logistic distribution under electronic commerce environment, the improved two-phase algorithm needs to be adopted to get solutions. Namely, the customer group can be divided into several regions using hierarchy clustering method in first phase. And in every region it can be decomposed into small scale subsets according with some restraint conditions using scan algorithm. In second phase, get the solutions of the customer point in every group using the improved genetic algorithm. In fact it is route optimization problems of several single TSP model. In the end, the test proves the validity of this improved algorithm combining with examples

Research on the Economic Features and Pricing of Digital Products

The expansion of Internet makes it a major channel and media to sell digital products. The distinctive economic features of digital products pull down the basis of traditional pricing strategies. We thus have to find a new pricing model suitable for information products in the network economy environment. When reasonable pricing strategies are applied, they must consider the benefit of both manufacturers and consumers. This article analyzes the pricing of digital products from horizontal differentiation and vertical differentiation, and provides relevant sales strategies

Research on the interactive effect of media attention and trading volume on stock return

Recent researchers have found stable relationship between media attention and stock price, trading volume as well as liquidity, but rare study focus on the interactive effect of media attention and trading volume on stock return. Our research bases on the listed bank stocks in China, uses the Baidu index data, and drive a panel data analysis. We find that there is a negative relationship between media attention and stock return, and a significant interactive effect of media attention and trading volume on stock return, that is, the effect of media attention on stock return increases when the trading volume increase. In the future, we will study the differences between countries to find more information about the effect, which can also enrich this research.

Study on Incentive Mechanism Model of Tacit Knowledge Sharing in Enterprises

Tacit knowledge exists in knowledge workers in enterprises. To enable full sharing, reasonable knowledge sharing incentive and restraint mechanisms should be designed to accelerate spontaneous behaviors of employees so that it can improve the ability of innovation and competitiveness. According to principal- agency theory, and taking into account incomplete information of knowledge sharing, multi-stages incentive model of knowledge sharing for knowledge worker, and incentive model of knowledge sharing for knowledge team based on team sharing interests are proposed. The results indicate that, in order to acquire maximum rewards in enterprise, special incentive contracts are to be made for different knowledge main-bodies, optimal incentive decisions to meet enterprise own objective are to be made, and knowledge body’s knowledge sharing effort is to be improved.

Study on the incentive mechanism model of tacit knowledge sharing

Tacit knowledge exists in knowledge workers in enterprises. To enable full sharing and improve core competitiveness, incentive mechanism models of knowledge sharing should be established for knowledge workers; According to principal- agency theory, and taking into account incomplete information of knowledge sharing, multi-stages incentive model of knowledge sharing for knowledge worker, incentive model of knowledge sharing for knowledge team based on team profits-sharing, and incentive model of knowledge sharing for the supplier, are proposed. The results indicate that, in order to acquire maximum rewards in enterprise, special incentive contracts are to be made for different knowledge main-bodies, optimal incentive decisions to meet enterprise own objective are to be made, and knowledge main-bodys knowledge sharing efforts are to be improved.

Analysis and design of the early-warning satellite scheduling simulation system

The early-warning satellite optimal scheduling is an effective way to improve its combat capability and performances. This paper analyzed first the early-warning satellite combat-flow and established a Petri Net model. Based on its particular combat pattern, this paper designed a simulation system of the early-warning satellite scheduling; described its architecture, function and data stream; discussed its key technologies and implementation. The simulation system was realized by the comprehensive application of MATLAB and STK.