碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / With the expending of Internet popularity, the competition between Internet Service Provider companies is achieving a peak. Because their products and services have highly similarity and also have close ranged pricing strategy. Hence, once ISP disappointed their customers, they will easily lose their market share. The purpose of this research is to understand the effect on different switch driver on customer’s emotion when they performed switching behavior. Second, after customer switched, what behavior can be influenced by customer’s emotion? This research combined Push-Pull-Mooring model and Expectation Disconfirmation Theory as drivers and post-switching problems. Moreover, the concept of Regret is used to investigate and explains the effect on customer’s behavior intention. The results of this research include: (1) Customers driven by Pull effects will cause higher Regret than driven by Push effects. (2) After switched, customers will sensed more Regret when negative consequences occurred. (3) Regret emotion created by post-switching behavior can affect customer’s negative behavior intention. The contribution of this research is to understand the causes of regret and the influences on customer’s post-switching behavior.

碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / Recently, the explosive development of mobile device has dramatically changed human life, mobile application becomes pervasive as well. Nowadays, there have released about 3 million mobile applications. Due to the tremendous and still increasing number of mobile application, user get harder to find needed apps. To tackle this problem, we propose a personalized recommender system based on the features of textual data. Specifically, we apply LDA to extract hidden topics of user reviews and use the probability of topic distribution to represent the features of app. Further, construct user profile based on his or her consumed apps in order to realize user requirements and preference. Eventually, for each app, we take account both the topic distribution and user preference to estimate a recommended score for target user and sort candidate apps by descending score to come out a personalized app recommended list. For the evaluation, we crawl the real-world dataset and adopt Recall@M as the measurement of performance. The experimental result shows our proposed mechanism outperforms the baseline and is able to enhance the performance of the state-of-the-art recommender systems. We then concluded that the user feedback is an effective variable to represent the features of app and plays significant role on app personalized recommender system.

碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / Competitor identification plays an important role in the success of a company. Companies have to identify their competitors to make the business strategies before confronting them. In this work, we propose a text mining and machine learning method to identify competitors for companies. Using the comparative opinions in customer reviews, we attempt to obtain the various information, including amount of co-occurrence and sentiment, associated with a pair of companies. Then, we build a classifier to identify competitors based on the co-occurrence and sentiment information. We finally demonstrate and evaluate our proposed method.

碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / The emergence of Web 2.0 has led to a new era of user generated content. More and more people tend to share their thinking, opinion and even user review of a specific product. Among all these textual data, not only useful paragraphs that are worth considering are in-cluded, there are also malicious comments and spams. Thus, how to determine the quality of text has become a serious problem, and a popular issue in the domain of text mining. Most of the existing studies use the characteristic of tokens as the analysis target, which cannot represent the comprehensive structure of a document, but only the combination of words, or statistical data such as the length of the document. Rhetorical Structure Theory (RST) transform a document to a tree that have the characteristic of level according to the relation between text spans. By incorporating the concept of RST, we can have a more accu-rate result on analyzing text quality. We proposed a process to analyze the quality of text, using NLTK (Natural Language Toolkit) to process the text, finding the potential predictor, and use different classification model to analyze real user reviews from Amazon.com. Also, an evaluation process is con-ducted to compare the result before and after adding RST features. Experiment shows that our model outperforms previous models which only consider the structure of word tokens.

碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / Due to the explosion of music tabs sharing websites, user can upload chord tabs with different key, structure, and format. As a result, there could be many tabs versions associated with one song, causing confusions to the users when it comes to choosing a tab. In order to solve this problem and ease labor cost to check, in this research we develop a similarity index that measures the similarity between a chord tab and its tonal progression of the original song, which may serve as the quality indicator of the tab. The proposed index requires the processing of chord tab, followed by Dynamic Time Warping(DTW) for computing the similarity between two chord sequences. Our experiment involves 1,633 chord tabs of 346 songs. We find that the proposed index values of chord tabs have positive correlation with the rank given by users (above 0.3) when comparing two chord sequences on the same key. We also find dynamic time warping (DTW) outperforms other similarity calculation methods, since DTW can compare two sequence varied in time and speed. The proposed similarity index can be used in determining the qualities of tabs as well as building chord tab recommendation system.

碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / Nowadays, the Internet provides diversified information. The enormous amounts of information such as online news are generated continuously as time goes by. The rapid-growth amount of online news makes it difficult to manually identify new and emerging events. Thereby, to solve this problem, we propose an approach using text mining techniques and topic modelling to detect the new events from broadcasting Chinese news sources automatically. To evaluate our method, we select our dataset from scrapping the Chinese news website of “AppleDaily” from 2008 to 2015, where each news articles of the corpus contains the keyword about Tourists from China. We use Chinese Natural Language Processing tool to preprocess our initial data. We implement Online-LDA topic model to find out new events. In the end, we conduct an experiment to measure the performance of our proposed method. The experimental results show that our proposed online event detection method is effective in detecting and tracking Chinese new events as news arrived in streams.

碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / The Contextual Bandit Problem (CMAB) is usually used to recommend for online applications on article, music, movie, etc. One leading algorithm for contextual bandit is the LinUCB algorithm, which updates internal linear regression models by the partial feedback from the environment. However, we observe that CMAB is rarely used in the stock recommendation, while most of the recommendations are for the purpose of profit, and ignore investor’s features (risk tolerance, investment features, and the others). We propose a personalized recommendation system for stock by using contextual multi-armed bandit algorithm. We take investor’s investment records as user features, and recommend the “arm”, which is a type of stock, based on two kinds of analysis, the technical and fundamental analysis. To the chosen arm, we rank the stocks according to the similarity of the stock and the arm. Our experiment is base on an online investment dataset, and the result demonstrates that our method outperforms other algorithms. Our experiment dataset collects simulation investment on the online website, and the result demonstrates that our method outperforms other algorithms.

碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / The emission volume increase of greenhouse gas (GHG) has caused the planet heating up and the climate change dramatically. No matter who causes the situation, such as society, enterprises or even individuals, all of them should take responsibilities in reducing the emission volume of the greenhouse gas. In recent years, the concept of the carbon footprint gradually rises. Consumers are more familiar to this concept and tend to choose the products that processes less carbon footprint. On the other hand, from the viewpoint of the enterprises, carbon footprint makes their supply chain more efficient and become more environmental friendly due to the concept of green supply chain. Our study employs ontology to build a model for the manufacturing processes in the real world. This model incorporates ontologies, statistic software and prolog language for inference, which can be used for carbon footprint reasoning. A prototype based on the proposed model is developed using Java and able to calculate the carbon footprint labeling of individual services or products. Finally, we evaluate our carbon footprint labeling approach and describe the research limitations and future works.

碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / Nowadays, people send a lot of information through social media. In addition, rumor is also an information type that spreads through people’s social media friend circles. Prior studies have focused on disastrous and catastrophic rumors that scared people, but many of today’s rumors are about our daily life. We want to investigate why people retransmit the rumor they hear and what factors will motivate them more to retransmit the rumor. To accomplish this, this study combines three theories to predict people’s retransmitting intention. The first is the Uses and Gratification Theory (U&G Theory). The second and third are the Rumor Retransmission Model and the Basic Law of Rumor. The former looks at people’s motivations to investigate what kinds of needs people want to fulfill and the latter seeks to understand what rumor characteristics will facilitate people to retransmit after receiving the rumor. The results of this study show that socializing, information seeking, and status seeking will increase the intent to retransmit; but entertainment has no significant effect. Personal involvement has no moderating effects on the relationships of socializing and information seeking to intention. Perceived credibility has a moderating effect on the relationship between socializing and intention, but is complementary with status seeking.

碩士 / 國立中山大學 / 資訊管理學系研究所 / 104 / As information technology changes with increasing frequency, industries have become heavily dependent upon a wide range of information systems (IS) to maintain and enhance organizational performance. Many companies have elected to outsource their information system projects. Over the years, scholars and practitioners alike have shown keen interest in the conditions that ensure the success of such outsourcing. This study explores the impact of information system project outsourcing from the perspective of the psychological contract breach. From this perspective, if the client feels that the obligations implied by the psychological contract are unfulfilled, or the company’s actions appear incongruent with the contract, this perceived breach in the psychological contract could lead to feelings of violation. This affective response causes the client to consider the outsourced IS system a failure. We conduct a questionnaire-based study, collecting 133 valid cases. Our respondents are IS department employees with experience in IS project outsourcing. Our results show that the psychological contract breach has significant influence on the success of IS project outsourcing. Moreover, in the context of IS outsourcing, the feelings of violation caused by this perceived breach of the contract can be attributed to reneging, incongruence and vigilance. Keywords: Psychological contract theory, IS outsourcing success, Psychological contract breach, Contracts, Outsourcing relationship