臺灣博碩士論文加值系統

隨著網際網路的普及，智慧型手機及平板電腦等行動裝置和應用程式的大量推出，造就了行動裝置作業系統(Android、iOS、Windows Phone)的一大商機，也為其軟體業者帶來廣大的市場與競爭力。其中遊戲類App為最多使用者使用及下載的類別，Juniper公司預測，在遊戲市場中，行動遊戲的收入到2020年將達到整體收入的3.5成(約360億美元)。在此競爭環境中，遊戲App開發商該如何設計出符合使用者需求且可以吸引使用者下載、持續使用並在App內進行消費，是值得關注的議題，特別是在目前遊戲App主要收益模式為遊戲內部購買(In-app purchases)的情況下。因此，本研究主要目的是以行動服務相關研究文獻來試圖定義影響遊戲App內部購買之潛在要素，再利用支持向量機遞歸特徵消除(SVM-RFE)、最小絕對壓縮挑選機制 (LASSO)、決策樹(DT)、神經網路修剪(Neural Network Pruning)與卡方檢定( test)特徵選取方法，以確認影響遊戲App內部購買的重要因素，研究結果希望能為遊戲開發商與研究學者做為參考之依據。

With the popularity of the Internet, mobile communications services and handheld devices, it led to explosive development in mobile applications (App). Related researchers also noted that game applications are the biggest source of revenue compared to other applications. New data from Juniper Research has found that the mobile games industry will account for ever increasing contributions, with a market share approaching 35% by 2020. Therefore, the game App providers need to know how to design products that match consumer’s needs, continued use and do in-app purchase are important issue. In particular, in-App purchases is the major revenue models. Hence, this study attempts to define the potential factors of influencing in-App purchases for game users. Then, we use five feature selection methods, Support Vector Machine-Recursive Feature Elimination (SVM-RFE), Least Absolute Shrinkage and Selection Operator (LASSO), Decision Tree (DT), Neural Network Pruning and Chi-square test ( test) to identify important factors that affect users in-App purchases behavior. The findings can be used as a reference when designing game Apps for game developers and researchers.

Abstract I
中文摘要 II
目錄 III
表目錄 V
圖目錄 VIII
第1章 緒論 1
1.1 研究背景 1
1.2 研究動機 6
1.3 研究目的 7
第2章 文獻探討 9
2.1 行動應用程式 9
2.2 遊戲App潛在要素定義 12
2.3 特徵選取(Feature Selection, FS) 16
2.3.1 支持向量機之遞歸特徵消除(Support Vector Machine-Recursive Feature, SVM-RFE) 19
2.3.2 最小絕對壓縮挑選機制 (Least absolute shrinkage and selection operator, LASSO) 20
2.3.3 決策樹(Decision Tree, DT) 21
2.3.4 神經網路修剪(Neural Network Pruning) 22
2.3.5 卡方檢定(Chi-square test, test) 23
第3章 研究方法 25
第4章 實驗結果與分析 35
4.1 衡量指標 35
4.2 要素定義 36
4.3 樣本敘述統計 38
4.4 特徵選取 41
4.4.1 支持向量機之遞歸特徵消除(SVM-RFE) 41
4.4.2 最小絕對壓縮挑選機制(LASSO) 43
4.4.3 決策樹(DT) 45
4.4.4 神經網路修剪(Neural Network Pruning) 46
4.4.5 卡方檢定( test) 50
4.5 SVM績效評估 53
第5章 結論與未來研究方向 59
5.1 結論 59
5.2 未來研究方向 68
參考文獻 69
附錄（問卷） 77

表目錄
表 2-1 遊戲 App 之相關研究文獻 11
表 2-2 行動遊戲因素文獻整理表（心理感知） 15
表 2-3 行動遊戲因素文獻整理表（遊戲品質） 15
表 4-1 影響行動遊戲內消費之要素定義表（心理感知） 38
表 4-2 影響行動遊戲內消費之要素定義表（遊戲品質） 39
表 4-3 受訪者基本資料統計 40
表 4-4 SVM-RFE 權重排名 43
表 4-5 SVM-RFE 因素選取 44
表 4-6 LASSO 因素選取 45
表 4-7 DT 因素選取 46
表 4-8 BPN 因素絕對乘法值 48
表 4-9 BPN 因素排名 49
表 4-10 BPN 因素選取 50
表 4-11 卡方因素排名 51
表 4-12 卡方因素選取 52
表 4-13 各特徵選取方法之顯著因素比較 53
表 4-14 SVM 評估 SVM-RFE 之特徵選取績效 54
表 4-15 SVM 評估 LASSO 之特徵選取績效 55
表 4-16 SVM 評估 DT 之特徵選取績效 55
表 4-17 SVM 評估 BPN 之特徵選取績效 56
表 4-18 SVM 評估卡方之特徵選取績效 57
表 4-19 SVM 評估五種特徵選取方法之績效 57
表 4-20 影響行動遊戲內消費之重要因素表 59
表 5-1 影響行動遊戲內消費之重要因素建議表 61
表 5-2 重要因素建議表(有消費經驗) 62
表 5-3 重要因素建議表(消費金額小於 150 元) 62
表 5-4 重要因素建議表(男性使用者) 63
表 5-5 重要因素建議表(女性使用者) 64
表 5-6 男女性使用者之重要因素比較 65
表 5-7 重要因素建議表(無消費經驗) 66
表 5-8 有無消費經驗之重要因素比較 67
表 5-9 重要因素建議表(消費金額大於 150 元) 68
表 5-10 消費金額大小之重要因素比較 68

圖目錄
圖 1-1 2008 年 7 月至 2017 年 1 月 App Store 中的 App 數量[84] 2
圖 3-1 研究流程圖 25
圖 4-1 特徵選取方法之績效比較 58

[1] 林芝儀 （2003），應用資料探勘於信用卡授信決策模式之實證研究，碩士論文，元智大學工業工程管理學系，台北。
[2] 陳隆昇、林瑞裕（2013），辨識科技產品操弄評論之研究，第七屆資訊科技國際研討會，台中。
[3] 陳隆昇、嚴振瑋（2015），確認影響下載手機遊戲App之關鍵因素，碩士論文，朝陽科技大學資訊管理系，台中。
[4] 楊正三、葉明龍、莊麗月、陳禹融、楊正宏（2008），「利用資訊增益與瀰集演算法於基因微陣列之特徵選取與分類問題」，資訊科技國際期刊，第二卷，第二期，第50-62頁。
[5] 楊正宏、柯兆軒、莊麗月、楊正三（2007），「用於基因微陣列資料的兩階段式基因選擇」，國際醫學資訊研討會論文集，花蓮，第16-17頁。
[6] 葉鎮源、楊維邦、蔡政容、林忠億（2015），「基於特徵關聯分群之特徵選取策略於資訊檢索排序模型學習的應用」，科技部補助專題研究計畫成果報告，編號：MOST 103-2221-E-178-001。
[7] 資策會，2015 https://mic.iii.org.tw/aisp/pressroom/press01_pop.asp?sno=400&type1=2
[8] 資策會，2016 https://mic.iii.org.tw/micnew/IndustryObservations_PressRelease02.aspx?sqno=423
[9] 劉博榮（2010），近體詩自動分類研究，碩士論文，國立交通大學資訊科學與工程研究所，新竹。
[10] 蕭宇翔（2005），應用MTS於非平衡資料分析之穩健性研究－以行動電話檢測流程為例，碩士論文，國立交通大學工業工程與管理學系，新竹。
[11] 戴于珊（2010），應用監督式類神經網路於遙測影像分類，碩士論文，國立交通大學資訊科學與工程研究所，新竹。
[12] Algamal, Z.Y., Lee, M.H., 2015, “Penalized logistic regression with the adaptive lasso for gene selection in high-dimensional cancer classification, ” Expert Systems With Applications, vol. 42, pp. 9326-9332.
[13] App Annie, 2015, https://www.appannie.com/indexes/all-stores/rank/overall/?_ref=header
[14] Archibald, R., Fann, G., 2014, “Feature selection and classification of hyperspectral images with support vector machines,” Computers and Electrical Engineering, vol. 40, pp. 16-28.
[15] Arteaga, S.M., Gonzalez, V.M., Kurniawan, S., and Benavides, R.A., 2012, “Mobile games and design requirments to increase teenagers’ physical activity,” Pervasive and Mobile Computing, vol. 8, pp. 900-908.
[16] Baker, M. S. A., Rosmiza, B., 2014, “Technology acceptance and purchase intention towards movie mobile advertising amoung youth in malaysia,” Procedia- Social and Behavioral Sciences, vol. 130, pp. 558-567.
[17] Basuroy, S., Chatterjee, S., and Ravid, S.A., 2003, “How critical are critical reviews? The box office effects of film critics,” Journal of Marketing, vol. 67, pp. 103-117.
[18] Bhattacherjee, A., 2001, “An empirical analysis of the antecedents of electronic commerce service continuance,” Decision Support Systems, vol. 32, pp. 201-214.
[19] Brims, F.J.H., Meniawy, T.M., Duffus, I., Fonseka, D.D., Segal, A., Creaney, J., Maskell, N., Lake, R.A., Klerk, N.D., and Nowak, A.K., 2016, “A novel clinical prediction model for prognosis in malignant pleural mesothelioma using decision tree analysis,” Journal of Thoracic Oncology, vol. 11, pp. 573-582.
[20] Cao, H., Lin, M., 2017, “Mining smartphone data for app usage prediction and recommendations: A survey,” Pervasive and Mobile Computing, vol. 37, pp. 1-22.
[21] Chan, N.H., Yau, C.Y., and Zhang, R.M., 2015, “Lasso estimation of threshold autoregressive models,” Journal of Econometrics, vol. 189, pp. 285-296.
[22] Chandrashekar, G., Sahin, F., 2014, “A survey on feature selection methods,” Computers and Electrical Engineering, vol. 40, pp. 16-28.
[23] Chen, C., Leung, L., 2015, “Are you addicted to candy crush saga? An exploratory study linking psychological factors to mobile social game addiction,” Telematics and Informatics, vol. 33 pp. 1155-1166.
[24] Chen, L.-S., Liu, C.-H. and Chiu, H.-J., 2011, “A neural network based approach for sentiment classification in the blogosphere,” Journal of Informetrics, vol. 5, pp. 313-322.
[25] Cnnic, 2014, The 33rd statistical report on the Internet Development in China, Retrieved from.
[26] ComScore, 2015, http://www.comscore.com/Insights/Presentations-and-Whitepapers/2015/The-2015-US-Mobile-App-Report
[27] Cota, T.T., Ishitani, L., and Jr, N.V., 2015, “Mobile game design for the elderly: a study with focus on the motivation to play,” Computers in Human Behavior, vol. 51, pp. 96-105.
[28] Davis, R., Lang, B., 2012, “Modeling game usage, purchase behavior and ease of use,” Entertainment Computing, vol. 3, pp. 27-36.
[29] Debska, B., Guzowska-Swider, B., 2011, “Decision trees in selection of featured determined food quality,” Analytica Chimica Acta, vol. 705, pp. 261-271.
[30] Distimo, 2014, http://www.distimo.com/
[31] Du, Q., Chang, C.I., 2001, “A linear constrained distance-based discriminant analysis for hyperspectral image classification,” Pattern Recognition, vol. 34, pp. 361-373.
[32] eMarketer, 2013, http://www.emarketer.com/Article/In-App-Purchases-Take-Over-App-Revenues/1010491
[33] eMarketer, 2014, https://www.emarketer.com/Article/Smartphone-Users-Worldwide-Will-Total-175-Billion-2014/1010536
[34] Ericsson Mobility Report, 2014, https://www.ericsson.com/news/141118-90percent-to-have-mobile-phones-by-2020-predicts-ericsson-mobility-report_244099435_c
[35] Erkaymaz, O., Ozer M., 2016, “Impact of small-world network topology on the conventional artificial neural network for the diagnosis of diabetes,” Chaos Soliton Fract, vol. 83, pp. 178-185.
[36] Feijoo, C., Gomez-Barroso, J.L., Aguado, J.M., and Ramos, S., 2012, “Mobile gaming: Industry challenges and policy implications,” Telecommunications Policy, vol. 36, pp. 212-221.
[37] Flurry, 2015, http://flurrymobile.tumblr.com/post/127638842745/seven-years-into-the-mobile-revolution-content-is
[38] Gao, L., Waechter, K.A., and Bai, X., 2015, “Understanding consumer’s continuance intention towards mobile purchase: A theoretical framework and empirical study- A case of China,” Computers in Human Behavior, vol. 53, pp. 249-262.
[39] Gharaei-Manesh, S., Fathzadeh, A., and Taghizadeh-Mehrjardi, R., 2016, “Comparison of artificial neural network and decision tree models in estimating spatial distribution of snow depth in a semi-arid region of Iran,” Cold Regions Science and Technology, vol. 122, pp. 26-35.
[40] Gheyas, I.A., Smith, L.S., 2010, “Feature subset selection in large simensionality domains,” Pattern Recognition, vol. 43, pp. 5-13.
[41] Guyon, I., Elisseeff, A., 2003, “An introduction to variable and feature selection,” Journal of Machine Learning Research, vol. 3, pp. 1157-1182.
[42] Guyon, I., Weston, J., Barnhill, S., Vapnik, V., 2002, “Gene selection for cancer classification using support vector machines,” Machine Learning, vol. 46, pp. 389-422.
[43] Ha, I., Yoon, Y., and Choi, M., 2007, “Determinants of adoption of mobile games under mobile broadband wireless access environment,” Information and Management, vol. 44, pp. 276-286.
[44] Hidalgo-Munoz, A.R., Lopez, M.M., Santos, I.M., Pereira, A.T., Vazquez-Marrufo, M., Galvao-Carmona, A., and Tome, A.M.P., 2013, “Application of svm-rfe on eeg signals for detecting the most relevant scalp regions linked to affective valence processing,” Expert Systems with Applications, vol. 40, pp. 2102-2108.
[45] Hsiao, C.H., Chang, J.J., and Tang, K.Y., 2016, “Exploring the influential factors in continuance usage of mobile social apps: Satisfaction, habit, and customer value perspectives,” Telematics and Informatics, vol. 33, pp. 342-355.
[46] Hsu, C.L., Lin, J.C.C., 2015, “What drives purchase intention for paid mobile apps? An expectation confirmation model with perceived value,” Electronic Commerce Research and Applications, vol. 14, pp. 46-57.
[47] Hsu, J.S.C., Lin, T.C., Fu, T.W., and Hung, Y.W., 2015, “The effect of unexpected features on app users’ continuance intention,” Electronic Commerce Research and Applications, vol. 14, pp. 418-430.
[48] Jansen, S., Bloemendal, E., 2013, “Defining app stores: The role of curated marketplaces in software ecosystems software business,” From Physical Products to Software Services and Solutions, Springer, vol. 150, pp. 195-206.
[49] Jun, S., 2016, “The reciprocal longitudinal relationships between mobile phone addiction and depressive symptoms among korean adolescents,” Computers in Human Behavior, vol. 58, pp. 179-186.
[50] Juniper, 2015, http://www.juniperresearch.com/press/press-releases/digital-games-revenues-to-pass-$100-billion-per-an
[51] Kang, J.Y.M., Mun, J.M., and Johnson, K.K.P, 2015, “In-store mobile usage: Downloading and usage intention toward mobile location-based retail apps,” Computers in Human Behavior, vol. 46, pp. 210-217.
[52] Kim, B., 2010, “An empirical investigation of mobile data service continuance: Incorporating the theory of planned behavior into the expection-confirmation model,” Expert Systems with Applications, vol. 37, pp. 7033-7039.
[53] Kubat, M., Matwin, S., 1997, “Addressing the curse of imbalanced data sets: one-sided sampling,” in 14th International conference on Machine Learning, pp. 179-186.
[54] Kwon, S., Lee, S., and Kim, Y., 2015, “Moderately clipped lasso,” Somputational Statistics and Data Analysis, vol. 92, pp. 53-67.
[55] Lazar, C., Taminau, J., Meganck, S., Steenhoff, D., Coletta, A., Molter, C., Schaetzen, V.D., Duque, R., Bersini, H., and Nowe, A., 2012, “A survey on filter techniques for feature selection in gene expression microarray analysis,” IEEE/ACM Transactions on Computational Biology and Bioinformatics, vol. 9, pp. 1106-1119.
[56] Lee, A., Song, K., Ryu, H.B., Kim, J., and Kwon, G., 2015a, “Fingerstroke time estimates for touchscreen-based mobile gaming interaction,” Human Movement Science, vol. 44, pp. 211-224.
[57] Lee, J., Lee, J., Lee, H., and Lee, J., 2015b, “An exploratory study of factors influencing repurchase behaviors toward game items: A field study,” Computers in Human Behavior, vol. 43, pp. 13-23.
[58] Li, T., Zhang, C., Ogihara, M., 2004, “A comparative study of feature selection and multiclass classification methods for tissue classification based on gene expression,” Science and Mathematics, vol. 20, pp. 2429-2437.
[59] Li, X.-B., Qiu, J.-K., Han, S., 2016, “Feature selection method based on linear twin support vector machine,” Computer Engineering and Applications, vol. 52 pp. 88-92.
[60] Lin, H.H., Wang, Y.S., 2006, “An examination of the determinants of customer loyalty in mobile commerce contexts,” Information and Management, vol. 43, pp. 271-282.
[61] Ma, L., Destercke, S., and Wang, Y., 2016, “Online active learning of decision trees with evidential data,” Pattern Recognition, vol. 52, pp. 33-45.
[62] McMillan, B., Hickey, E., Patel, M.G., and Mitchell, G., 2016, “Quality assessment of a sample of mobile app-based health behavior change interventions using a tool based on the national institute of health and care excellence behavior change guidance,” Patient Education and Counseling, vol. 99, pp. 429-435.
[63] Mic, 2012, http://mic.iii.org.tw/micnew/IndustryObservations_PressRelease02.aspx?sqno=310
[64] Mobilewalla, 2011, https://www.mobilewalla.com/
[65] Moutis, P., Skarvelis-Kazakos, S., and Brucoli, M., 2016, “Decision tree aided planning and energy balancing of planned community microgrids,” Applied Energy, vol. 161, pp. 197-205.
[66] Myers, D., 1990, “Computer games genres,” Play and Culture, vol. 3, pp. 286-301.
[67] Park, E., Baek, S., Ohm, J., and Chang, H.J., 2014, “Determinants of player acceptance of mobile social network games: An application of extended technology acceptance model,” Teiematics and Informatics, vol. 31, pp. 3-15.
[68] Park, H.J., Kim, S.H., 2013, “A bayesian network approach to examining key success factors of mobile games,” Journal of Business Research, vol. 66, pp. 1353-1359.
[69] Peng, H., Long, F., and Ding, C., 2005, “Feature selection based on mutual information: Criteria of max-dependency, max-relevance, and min-redundancy,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 27, pp. 1226-1238.
[70] Rakkiyappan, R., Velmurugan G. and Cao, J.D., 2015, “Stability analysis of fractional-order complex-valued neural networks with time delays,” Chaos Soliton Fract, vol. 78, pp. 297-316.
[71] Recker, J., 2010, “Explaining usage of process modeling grammars: Comparing three theoretical models in the study of two grammars,” Information and Management, vol. 47, pp. 316-324.
[72] Reed, R., 1993, Pruning algorithms, IEEE Transaction on Neural Networks, vol. 5, pp. 740–747.
[73] Rice, M., Koh, R.Y.I., and Ng, J., 2016, “Investigating gesture-based avatar game representations in teenagers, younger and older adults,” Entertainment Computing, vol. 12, pp. 40-50.
[74] Saeys, Y., Inza, I., and Larranaga, P., 2007, “A review of feature selection techniques in bioinformatics,” Science and Mathematics, vol. 23, pp. 2507-2517.
[75] Shu, W., Shen, H., 2016, “Multi-criteria feature selection on cost-sensitive data with missing values,” Pattern Recognition, vol. 51, pp. 268-280.
[76] Sok, H.K., Ooi, M.P.L., Kuang, Y.C., and Demidenko, S., 2016, “Multivariate alternating decision trees,” Pattern Recognition, vol. 50, pp. 195-209.
[77] Spetable, F.E., Bulacio, P., Guillaume, S., Murillo, J., and Tapia, E., 2015, “A spectral envelope approach towards effective svm-rfe on infrared data,” Pattern Recognition Letters, vol. 71, pp. 59-65.
[78] Su, C.-T., Chen, L.-S., & Chiang, T.-L., 2006, A neural network based information granulation approach to shorten the cellular phone test process, Computers in Industry, vol. 57, pp. 412–423.
[79] Su, C.-T., Hsu, J.-H., and Tsai, C.-H., 2002, Knowledge mining from trained neural network, Journal of Computer Information Systems, pp. 61–70.
[80] Sun, W., Xu, Y., 2016, “Financial security evaluation of the electric power industry in China based on a back propagation neural network optimized by genetic algorithm,” Energy, vol. 101, pp. 366-379.
[81] Sweeney, J.C., Soutar, G.N., 2001, “Consumer perceived value: The development of a multiple item scale,” Journal of Retailing, vol. 77, pp. 203-220.
[82] Tan, P.N., Steinbach, M., and Kumar, V., 2006, “Introduction to data mining,” Posts and Telecom Press.
[83] Thaseen, I. S., Kumar, C. A., 2016, “Intrusion detection model using fusion of chi-square feature selection and multi class SVM,” Journal of King Saud University - Computer and Information Sciences, In Press, Corrected Proof.
[84] The Statistics Portal, 2017, https://www.statista.com/statistics/263795/number-of-available-apps-in-the-apple-app-store/
[85] Thimm, G., Fiesler, E., 1997, Pruning of neural networks (IDIAIP-RR 97-03). Martigny, Valais: Switzerland. Dalle Molle Institute for Perceptive Artificial Intelligence.
[86] Thong, J.Y.L., Hong, S.J., and Tam, K.Y., 2006, “The effects of post-adoption beliefs on the expectation-confirmation model for information technology continuance,” International Journal of Human and Computer Studies, vol. 64, pp. 799-810.
[87] Tibshirani, R., 1996, “Regression shrinkage and selection via the lasso,” Journal of the Royal Statistical Society, Series B, vol. 58, pp. 267-288.
[88] Vinh, N.X., Zhou, S., Chan, J., and Bailey, J., 2016, “Can high-order dependencies improve mutual information based feature selection? ” Pattern Recognition, vol. 53, pp. 46-58.
[89] Wang, S., Chang, X., Li, X., Sheng, Q.Z., and Chen, W., 2016, “Multi-task support vector machines for feature selection with shared knowledge discovery,” Signal Processing, vol. 120, pp. 746-753.
[90] Wang, X., Chen, H., Chang, C., Jiang, M., Wang, X. and Xu, L., 2017, “Study the therapeutic mechanism of amomum compactum in gentamicin-induced acute kidney injury rat based on a back propagation neural network algorithm,” Journal of Chromatography B, vol. 1040, pp. 1-88.
[91] Wei, L., Liao, M., Gao, Y, Ji, R., He, Z. and Zou, Q., 2014, “Improved and promising identification of human micrornas by incorporating a high-quality negative set,” IEEE/ACM Transactions on Computational Biology and Bioinformatics, vol. 11, pp. 192-201.
[92] West, J., Mace, M., 2010, “Browsing as the killer app: Explaining the rapid success of apple's iphone,” Telecommunications Policy, vol. 34, pp. 270-286.
[93] Yan, K., Zhang, D., 2015, “Feature selection and analysis on correlated gas sensor data with recursive feature elimination,” Sensors and Actuators B: Chemical, vol. 212, pp. 353-363.
[94] Yan, Z., Yao, Y., 2015, “Variable selection method for fault isolation using least absolute shrinkage and selection operator (lasso),” Chemometrics and Intelligent Laboratory Systems, vol. 146, pp. 136-146.
[95] Yang, C.C., 2011, “A classification-based kansei engineering system for modeling consumers' affective responses and analyzing product form features,” Expert Systems with Applications, vol. 38, pp. 11382-11393.
[96] Yang, D., Ye, Z., Lim, L.H.I., and Dong, Z., 2015, “Very short term irradiance forecasting using the lasso,” Solar Energy, vol. 114, pp. 314-326.
[97] Yoon, S., Kim, S., 2009, “Mutual information-based svm-rfe for diagnostic classification of digitized mammograms,” Pattern Recognition Letters, vol. 30, pp. 1489-1495.
[98] Zhou, Q., Zhou, H., and Li, T., 2016, “Cost-sensitive feature selection using random forest: Selecting low-cost subsets of informative features,” Knowledge-Based System, vol. 95, pp. 1-11.
[99] Zhou, Q., Zhou, H., Yang, F., and Luo, L., 2014, “Structure damage detection based on random forest recursive feature elimination,” Mechanical Systems and Signal Processing, vol. 46, pp. 82-90.