Browsing by Author "Liu, Duen-Ren"

Now showing 1 - 20 of 75

A Hybrid Neural Network Model for Sales Forecasting Based on ARIMA and Search Popularity of Article Titles
(2016-01-01)
Omar, Hani
;
Van Hai Hoang
;
Liu, Duen-Ren
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
Enhancing sales and operations planning through forecasting analysis and business intelligence is demanded in many industries and enterprises. Publishing industries usually pick attractive titles and headlines for their stories to increase sales, since popular article titles and headlines can attract readers to buy magazines. In this paper, information retrieval techniques are adopted to extract words from article titles. The popularity measures of article titles are then analyzed by using the search indexes obtained from Google search engine. Backpropagation Neural Networks (BPNNs) have successfully been used to develop prediction models for sales forecasting. In this study, we propose a novel hybrid neural network model for sales forecasting based on the prediction result of time series forecasting and the popularity of article titles. The proposed model uses the historical sales data, popularity of article titles, and the prediction result of a time series, Autoregressive Integrated Moving Average (ARIMA) forecasting method to learn a BPNN-based forecasting model. Our proposed forecasting model is experimentally evaluated by comparing with conventional sales prediction techniques. The experimental result shows that our proposed forecasting method outperforms conventional techniques which do not consider the popularity of title words.
A hybrid of sequential rules and collaborative filtering for product recommendation
(2007)
Liu, Duen-Ren
;
Lai, Chin-Hui
;
Lee, Wang-Jung
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
Customers' purchase behavior may vary over time. Traditional collaborative filtering (CF) methods make recommendations to a target customer based on the purchase behavior of customers whose preferences are similar to those of the target customer,- however, the methods do not consider how the customers' purchase behavior may vary over time. Although the sequential rule method considers the sequence of customers' purchase behavior over time, it does not make use of the target customer's purchase data for the current period. To resolve the above problems, this work proposes a novel hybrid recommendation method that combines the segmentation-based sequential rule method with the segmentation-based CF method. Experiment. results show that the hybrid method outperforms traditional CF methods. Keywords:
A hybrid of sequential rules and collaborative filtering for product recommendation
(2009-09-29)
Liu, Duen-Ren
;
Lai, Chin-Hui
;
Lee, Wang-Jung
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
Customers' purchase behavior may vary over time. Traditional collaborative filtering (CF) methods make recommendations to a target customer based on the purchase behavior of customers whose preferences are similar to those of the target customer; however, the methods do not consider how the customers' purchase behavior may vary over time. In contrast, the sequential rule-based recommendation method analyzes customers' purchase behavior over time to extract sequential rules in the form: purchase behavior in previous periods double right arrow purchase behavior in the current period. If a target customer's purchase behavior history is similar to the conditional part of the rule, then his/her purchase behavior in the current period is deemed to be the consequent part of the rule. Although the sequential rule method considers the sequence of customers' purchase behavior over time, it does not utilize the target customer's purchase data for the current period. To resolve the above problems, this work proposes a novel hybrid recommendation method that combines the segmentation-based sequential rule method with the segmentation-based KNN-CF method. The proposed method uses customers' RFM (Recency, Frequency, and Monetary) values to cluster customers into groups with similar RFM values. For each group of customers, sequential rules are extracted from the purchase sequences of that group to make recommendations. Meanwhile, the segmentation-based KNN-CF method provides recommendations based on the target customer's purchase data for the current period. Then, the results of the two methods are combined to make final recommendations. Experiment results show that the hybrid method outperforms traditional CF methods. (C) 2009 Elsevier Inc. All rights reserved.
A LOGICAL FORMULATION OF ROUGH SET DEFINABILITY
(2011-02-01)
Fan, Tuan-Fang
;
Liau, Churn-Jung
;
Liu, Duen-Ren
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
In data mining problems, data is usually provided in the form of data tables. To represent knowledge discovered from data tables, a decision logic (DL) is proposed in rough set theory. DL, is an instance of propositional logic, but we can use other logical formalisms to describe data tables. In this paper, we propose two descriptions of data tables based on first-order data logic (FOUL) and attribute value-sorted logic (AVSL) respectively. In the context of FODL, we show that explicit, definability and implicit definability in classical logic implies the notion of definability in rough set theory. We also show that AVSL is particularly useful for the representation of properties of many-valued data tables.
A relational perspective of attribute reduction in rough set-based data analysis
(2011-08-16)
Fan, Tuan-Fang
;
Liau, Churn-Jung
;
Liu, Duen-Ren
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
Attribute reduction is very important in rough set-based data analysis (RSDA) because it can be used to simplify the induced decision rules without reducing the classification accuracy. The notion of reduct plays a key role in rough set-based attribute reduction. In rough set theory, a reduct is generally defined as a minimal subset of attributes that can classify the same domain of objects as unambiguously as the original set of attributes. Nevertheless, from a relational perspective, RSDA relies on a kind of dependency principle. That is, the relationship between the class labels of a pair of objects depends on component-wise comparison of their condition attributes. The larger the number of condition attributes compared, the greater the probability that the dependency will hold. Thus, elimination of condition attributes may cause more object pairs to violate the dependency principle. Based on this observation, a reduct can be defined alternatively as a minimal subset of attributes that does not increase the number of objects violating the dependency principle. While the alternative definition coincides with the original one in ordinary RSDA, it is more easily generalized to cases of fuzzy RSDA and relational data analysis. (C) 2010 Elsevier B.V. All rights reserved.
A Uniform Framework for Rough Approximations Based on Generalized Quantifiers
(2015-01-01)
Fan, Tuan-Fang
;
Liau, Churn-Jung
;
Liu, Duen-Ren
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
The rough set theory provides an effective tool for decision analysis in the way of extracting decision rules from information systems. The rule induction process is based on the definitions of lower and upper approximations of the decision class. The condition attributes of the information system constitute an indiscernibility relation on the universe of objects. An object is in the lower approximation of the decision class if all objects indiscernible with it are in the decision class and it is in the upper approximation of the decision class if some objects indiscernible with it are in the decision class. Various generalizations of rough set theory have been proposed to enhance the capability of the theory. For example, variable precision rough set theory is used to improve the robustness of rough set analysis and fuzzy rough set approach is proposed to deal with vague information. In this paper, we present a uniform framework for different variants of rough set theory by using generalized quantifiers. In the framework, the lower and upper approximations of classical rough set theory are defined with universal and existential quantifiers respectively, whereas variable precision rough approximations correspond to probability quantifiers. Moreover, fuzzy rough set approximations can be defined by using different fuzzy quantifiers. We show that the framework can enhance the expressive power of the decision rules induced by rough set-based decision analysis.
Advertisement recommendation based on personal interests and ad push fairness
(2019-01-01)
Liu, Duen-Ren
;
Liao, Yu-Shan
;
Chung, Ya-Han
;
Chen, Kuan-Yu
;
資訊管理與財務金融系註：原資管所+財金所
Purpose - Online advertisement brings huge revenue to many websites. There are many types of online advertisement; this paper aims to focus on the online banner ads which are usually placed in a particular news website. The investigated news website adopts a pay-per-ad payment model, where the advertisers are charged when they rent a banner from the website during a particular period. In this payment model, the website needs to ensure that the ad pushed frequency of each ad on the banner is similar. Under such advertisement push rules, an ad-recommendation mechanism considering ad push fairness is required. Design/methodology/approach - The authors proposed a novel ad recommendation method that considers both ad-push fairness and personal interests. The authors take every ad's exposure time into consideration and investigate users' three different usage experiences in the website to identify the main factors affecting the interests of users. Online ad recommendation is conducted on the investigated news website. Findings - The results of the experiments show that the proposed approach performs better than the traditional approach. This method can not only enhance the average click rate of all ads in the website but also ensure reasonable fairness of exposure frequency of each ad. The online experiment results demonstrate the effectiveness of this approach. Originality/value - Existing researches had not considered both the advertisement recommendation and ad-push fairness together. With the proposed novel ad recommendation model, the authors can improve the ad click-through rate of ads with reasonable push fairness. The website provider can thereby increase the commercial value of advertising and user satisfaction.
Air pollution forecasting based on attention-based LSTM neural network and ensemble learning
(1970-01-01)
Liu, Duen-Ren
;
Lee, Shin-Jye
;
Huang, Yang
;
Chiu, Chien-Ju
;
科技管理研究所
With air pollution having become a global concern, scientists are committed to working on its amelioration. In the field of air pollution prediction, there have been good results in experimental research so far, but few studies have integrated weather forecast information and the properties of air pollution drift. In this work, we propose a novel wind-sensitive attention mechanism with a long short-term memory (LSTM) neural network model to predict the air pollution - PM2.5 concentrations by considering the influence of wind direction and speed on the changes of spatial-temporal PM2.5 concentrations in neighbouring areas. Preliminary predictions for PM2.5 are then made by an LSTM neural network regarding neighbouring pollution; these predictions are "paid attention to" and we finally apply an ensemble learning method based on eXtreme Gradient Boosting (XGBoost) to combine the preliminary predictions with weather forecasting to make second phase predictions of PM2.5. The experiment is conducted using PM2.5 data and weather forecast data. Our results illustrate that the proposed method is superior to other methods in predicting PM2.5 concentrations, including multi-layer perceptron, support vector regression, LSTM neural network, and extreme gradient boosting algorithm.
An Online Activity Recommendation Approach based on the Dynamic Adjustment of Recommendation Lists
(2017-01-01)
Liu, Duen-Ren
;
Chen, Kuan-Yu
;
Chou, Yun-Cheng
;
Lee, Jia-Huei
;
資訊管理與財務金融系註：原資管所+財金所
This research investigates an online recommendation method for new types of online news websites. Cross-domain analysis on user browsing news and the attending activities is conducted to predict user preferences on activities based on non-negative Matrix Factorization (NMF) and Latent Dirichlet Allocation (LDA) topic model. A novel approach is proposed for the dynamic adjustment of recommendation lists in order to tackle the issue of limited recommendation layouts. The existing studies have not addressed this issue. The proposed approach is implemented on an online news website and evaluated for online recommendations. The experiment results demonstrate that our method can predict user preferences on recommended activities and enhance the effectiveness of recommendations.
Arrow decision logic for relational information systems
(2006)
Fan, Tuan-Fang
;
Liu, Duen-Ren
;
Tzeng, Gwo-Hshiung
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
In this paper, we propose an arrow decision logic (ADL) for relational information systems (RIS). The logic combines the main features of decision logic (DL) and arrow logic (AL). DL represents and reasons about knowledge extracted from decision tables based on rough set theory, whereas AL is the basic modal logic of arrows. The semantic models of DL are functional information systems (FIS). ADL formulas, on the other hand, are interpreted in RIS. RIS, which not only specifies the properties of objects, but also the relationships between objects. We present a complete axiomatization of ADL and discuss its application to knowledge representation in multicriteria decision analysis.
Churn prediction and social neighbour influences for different types of user groups in virtual worlds
(2019-06-01)
Liu, Duen-Ren
;
Liao, Hsiu-Yu
;
Chen, Kuan-Yu
;
Chiu, Yi-Ling
;
資訊管理與財務金融系註：原資管所+財金所
In virtual worlds (VWs), users have more VW games alternatives, whereas VW companies consequently suffer from high customer turnover rate and low customer loyalty. Therefore, building a churn prediction model to facilitate subsequent churn management and customer retention is important. The churn behaviours and the impact of social neighbour influences to customer churn may be different for different types of users. Accordingly, we segment users into stable, unstable, and solitary user groups according to their social contact behaviours in VWs. Novel segmentation-based churn prediction approaches are proposed for churn prediction in VWs by building prediction models for each type of user groups and considering the effect of social neighbour influences for different user groups. The proposed approaches are evaluated by conducting experiments with a dataset collected from a VW platform. The experimental results show different churn prediction performances under different user groups. The segmentation-based churn prediction approaches perform better than do general approaches without considering user groups. Moreover, the results also reveal that social neighbour influences have a positive impact on stable and unstable users. The proposed work contributes to investigating the social neighbour influences on churn prediction for different types of user groups in VWs.
Collaborative relevance assessment for task-based knowledge support
(2008-01-01)
Liu, Duen-Ren
;
Wu, I-Chin
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
The operations and management activities of enterprises are mainly task-based and knowledge intensive. Accordingly, an important issue in deploying knowledge management systems is the provision of task-relevant information (codified knowledge) to meet the information needs of knowledge workers during the execution of a task. Codified knowledge extracted from previously executed tasks can provide valuable knowledge about conducting the task-at-hand (current task), and is a valuable information source for constructing a task profile that models a worker's task needs, i.e., information needs for the current task. In this paper, we propose a novel task-relevance assessment approach that evaluates the relevance of previous tasks in order to construct a task profile for the current task. The approach helps knowledge workers assess the relevance of previous tasks through linguistic evaluation and the collaboration of knowledge workers. In addition, applying relevance assessment to a large number of tasks may create an excessive burden for workers. Thus, we propose a novel two-phase relevance assessment method to help workers conduct relevance assessment effectively. Furthermore, a modified relevance feedback technique, which is integrated with the task-relevance assessment method, is employed to derive the task profile for the task-at-hand. Consequently, task-based knowledge support can be enabled to provide knowledge workers with task-relevant information based on task profiles. Empirical experiments demonstrate that the proposed approach models workers' task-needs effectively and helps provide task-relevant knowledge. (C) 2007 Elsevier B.V. All rights reserved.
Combining Reputation and Content-Based Filtering for Blog Article Recommendation in Social Bookmarking Websites
(2010)
Peng, Chi-Chieh
;
Liu, Duen-Ren
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
The new generation of web-based communities, Web2.0, represents an innovative spirit in sharing and managing contents. Social bookmarking is a portal for users to share, organize, search, and manage bookmarks of web resources. However, with the rapid growth of web documents that are produced every day, people are facing the problem of information overload. The Social bookmarking web site provides the push (user recommendation) counts of articles indicating the recommended popularity degrees of articles. In this paper, we propose to derive the popularity degree of an article by considering the reputation of users that push the article. Moreover, we propose a personalized blog article recommendation approach, which combines the reputation-based popularity with content based filtering, to recommend popular blog articles to users that satisfy their personal preferences. Our experimental results show that the proposed approach outperforms conventional approaches.
Complementary QA Network Analysis for QA Retrieval in Social Question-Answering Websites
(2015-01-01)
Liu, Duen-Ren
;
Chen, Yu-Hsuan
;
Shen, Minxin
;
Lu, Pei-Jung
;
資訊管理與財務金融系註：原資管所+財金所
With the ubiquity of the Internet and the rapid development of Web 2.0 technology, social question and answering (SQA) websites have become popular knowledge-sharing platforms. As the number of posted questions and answers (QAs) continues to increase rapidly, the massive amount of question-answer knowledge is causing information overload. The problem is compounded by the growing number of redundant QAs. SQA websites such as Yahoo! Answers are open platforms where users can freely ask or answer questions. Users also may wish to learn more about the information provided in an answer so they can use related keywords in the answer to search for extended, complementary information. In this article, we propose a novel approach to identify complementary QAs (CQAs) of a target QA. We define two types of complementarity: partial complementarity and extended complementarity. First, we utilize a classification-based approach to predict complementary relationships between QAs based on three measures: question similarity, answer novelty, and answer correlation. Then we construct a CQA network based on the derived complementary relationships. In addition, we introduce a CQA network analysis technique that searches the QA network to find direct and indirect CQAs of the target QA. The results of experiments conducted on the data collected from Yahoo! Answers Taiwan show that the proposed approach can more effectively identify CQAs than can the conventional similarity-based method. Case and user study results also validate the helpfulness and the effectiveness of our approach.
Constraint-Based Attribute Reduction in Rough Set Analysis
(2010-01-01)
Fan, Tuan-Fang
;
Liau, Churn-Jung
;
Liu, Duen-Ren
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
Attribute reduction is very important in rough set-based data analysis (RSDA) because it can be used to simplify the induced decision rules without reducing the classification accuracy. The notion of reduct plays a key role in rough set-based attribute reduction. In rough set theory, a reduct is generally defined as a minimal subset of attributes that can classify the same domain of objects as unambiguously as the original set of attributes. Nevertheless, from a relational perspective, RSDA relies on a kind of dependency constraint. That is, the relationship between the class labels of a pair of objects depends on the componentwise comparison of their condition attributes. The larger the number of condition attributes compared, the greater the probability that the constraint will hold. Thus, elimination of condition attributes may cause more object pairs to violate the constraint. Based on this observation, a reduct can be defined alternatively as a minimal subset of attributes that does not increase the number of objects violating the constraint. While the alternative definition coincides with the original one in ordinary RSDA, it is more easily generalized to cases of fuzzy RSDA and relational data analysis.
CONTEXT-BASED KNOWLEDGE SUPPORT FOR PROBLEM-SOLVING BY RULE-INFERENCE AND CASE-BASED REASONING
(2011-07-01)
Ke, Chih-Kun
;
Liu, Duen-Ren
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
Problem-solving is an important process that enables corporations to create competitive business advantages. Traditionally, case-based reasoning techniques have been widely used to help workers solve problems. However, conventional approaches focus on identifying similar problems without exploring relevant context of problem situations. Situation features are generally occurred according to the context characteristics of problems. Moreover, situation features collected are usually partial or incomplete. Workers need to use knowledge inferred from relevant context information and previous problem-solving experience to clarify the causes and take appropriate action effectively. In this paper, we propose to use rule inference to infer possible situation features based on context information. Association rule mining is used to discover context-based inference rules from historical problem-solving logs. The discovered patterns identify frequent associations between context information and situation features, and therefore, can be used to infer more situation features. By considering the inferred situation features, case-based reasoning can then be employed to identify similar situations effectively. Moreover, we employ information retrieval techniques to extract context-based situation profiles to model workers' information needs when handling problem situations in certain context. Effective knowledge support can thus be facilitated by providing workers with situation-relevant information based on the profiles. We develop a prototype system to demonstrate the effectiveness of providing context-based relevant information and decision-making knowledge to help workers solve problems.
Context-based knowledge support for problem-solving by rule-inference and case-based reasoning
(2008)
Liu, Duen-Ren
;
Ke, Chih-Kun
;
Wu, Mei-Yu
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
Problem-solving is an important process that enables corporations to create competitive business advantages. Traditionally, case-based reasoning techniques have been widely used to help workers solve problems. However, conventional approaches focus on identifying similar problems without exploring relevant context of problem situations. Situation features are generally occurred according to the context characteristics of problem. Moreover, situation features collected are usually partial or incomplete. Workers need to use knowledge inferred from relevant context information and previous problem-solving experience to clarify the causes and take appropriate action effectively. In this paper, we propose to use rule inference to infer possible situation features based on context information. Association rule mining is used to discover context-based inference rules from historical problem-solving logs. The discovered patterns identify frequent associations between context information and situation features, and therefore can be used to infer more situation features. By considering the inferred situation features, case-based reasoning can then be employed to identify similar situations effectively. Moreover, we employ information retrieval techniques to extract context-based situation profiles to model workers' information needs when handling problem situations in certain context. Effective knowledge support can thus be facilitated by providing workers with situ ation-releva nt information based on the profiles.
Customer Churn Prediction in Virtual Worlds
(2015)
Liao, Hsiu-Yu
;
Chen, Luan-Yu
;
Liu, Duen-Ren
;
Chiu, Yi-Ling
;
資訊管理與財務金融系註：原資管所+財金所
With the emerging of social network websites, more and more social network online games are booming. Players have more alternatives of VWs games, while platform providers suffer from the problems of high customer turnover rate and low-customer-loyalty. Therefore, building a churn prediction model to facilitate subsequent churn management and customer retention is the best core marketing strategy. In this paper, we put emphasis on modeling a hybrid classification, which takes monetary cost, user behavior and social neighbor features into consideration. The experimental results show that the proposed hybrid model is well-suited for this problem in virtual worlds comparing to the existing churn prediction methods applied in the traditional retail, and financial industry.
Definability in Logic and Rough Set Theory
(2008-01-01)
Fan, Tuan-Fang
;
Liau, Churn-Jung
;
Liu, Duen-Ren
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
Rough set theory is an effective tool for data mining. According to the theory, a concept is definable if it can be written as a Boolean combination of equivalence classes induced from classification attributes. On the other hand, definability in logic has been explicated by Beth's theorem. In this paper, we propose two data representation formalisms, called first-order data logic (FODL) and attribute value-sorted logic (AVSL), respectively. Based on these logics, we explore the relationship between logical definability and rough set definability.
Discovering competitive intelligence by mining changes in patent trends
(2010-04-01)
Shih, Meng-Jung
;
Liu, Duen-Ren
;
Hsu, Ming-Li
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
Obtaining sufficient competitive intelligence is a critical factor in helping business managers gain and maintain competitive advantages. Patent data is an important source of competitive intelligence that enterprises can use to gain a strategic advantage. Under existing approaches, to detect changes in patent trends, business managers must rely on patent analysts to compare two patent analysis charts of different time periods. The discovery of change of trends currently still needs laborious human efforts and no efficient computer-based approaches are available for helping this task. In this paper, we propose a patent trend change mining (PTCM) approach that can identify changes in patent trends without the need for specialist knowledge. The proposed approach consists of steps including patent collection, patent indicator calculation, and change detection. In change detection phase the approach firstly excavate rules between two different time periods, comparing them to determine the trend changes. These trend changes are then classified into four categories of change, evaluated with change degree and ranked by their change degree as the output information to be referred by decision makers. We apply the PTCM approach to Taiwan's semiconductor industry to discover changes in four types of patent trends: the R&D activities of a company, the R&D activities of the industry, company activities in the industry and industry activities generally. The proposed approach generates competitive intelligence to help managers develop appropriate business strategies. (C) 2009 Elsevier Ltd. All rights reserved.