消費者對可追溯食品支付意愿建模及分析：潛類別模型及計算機模擬
發(fā)布時間：2019-03-12 10:15
【摘要】：近年來，國內(nèi)食品安全質(zhì)量問題十分突出，引發(fā)了人們對食品安全的強烈關注。食品可追溯體系通過在供應鏈上形成可靠且連續(xù)的安全信息流，能夠監(jiān)控食品生產(chǎn)過程與流向且通過追溯來識別問題源頭和實施召回，被認為是有效消除信息不對稱，從根本上預防食品安全風險的主要工具之一。本文對消費者對可追溯屬性支付意愿展開研究。針對國內(nèi)沒有大規(guī)模的食品安全調(diào)研數(shù)據(jù)，本文通過研究食品可追溯性應包含的可追溯信息與層次，運用菜單選擇實驗法設計問卷進行調(diào)研，將調(diào)研結果作為研究分析的數(shù)據(jù)輸入。由于菜單法數(shù)據(jù)是離散的分類屬性數(shù)據(jù)的特點，一般方法難以應用。針對以上問題本文綜合運用經(jīng)濟學的潛類別模型和適用于分類屬性數(shù)據(jù)的k-modes聚類算法研究消費者對豬肉可追溯屬性的支付意愿從而分析消費者的群體性偏好，以擴大消費者對可追溯食品的需求，更好的推廣食品可追溯體系，保障食品安全。本文主要工作如下： (1)通過對國內(nèi)外食品可追溯體系及支付意愿相關文獻研究與實地調(diào)研，設計豬肉供應鏈體系的可追溯信息應該包括養(yǎng)殖、屠宰加工和配送銷售及政府認證四個屬性。對上述四個可追溯屬性分別設置不同的價格層次設計菜單選擇實驗法問卷，以D-efficiency檢驗顯示問卷設計優(yōu)良。在江蘇省無錫市實地問卷調(diào)研收集數(shù)據(jù),問卷結果統(tǒng)計分析顯示調(diào)查結果良好。 (2)根據(jù)消費者效用理論針對消費者對可追溯支付意愿建立潛類別模型，以消費者的類別為潛在變量，以消費者的選擇作為外顯變量，運用菜單選擇實驗法調(diào)查數(shù)據(jù)對消費者行為分析。結果表明消費者對食品可追溯屬性的需求呈現(xiàn)出不同的偏好，普遍屬于低水平可追溯屬性消費群體。 (3)通過對聚類相關算法研究，針對菜單選擇實驗法數(shù)據(jù)是離散的分類屬性數(shù)據(jù)的特點，找到適合聚類分析的k-modes算法。針對k-modes算法存在的聚類過程復雜，分類精度不高的問題結合最新研究進展綜合改進k-modes聚類，通過結合密度和距離兩個因素選取初始聚類中心從而簡化聚類過程，以考慮可追溯屬性的所有屬性值的模式代替k-modes聚類算法的modes，從而提高聚類分類精確性。 (4)根據(jù)調(diào)查問卷建立聚類分析模型，將改進k-modes聚類方法應用于菜單選擇實驗法問卷結果分析，，以CU和目標函數(shù)走向來選取合適的聚類類別數(shù)目。研究結果表明，消費者可分為多個對可追溯安全信息偏好不同的群體，這些群體支付能力也不同�？舍槍Σ煌娜后w提供不同的可追溯屬性組合的豬肉以擴大消費者對可追溯食品的需求，提高食品安全保障水平。
[Abstract]:In recent years, the domestic food safety quality problem is very prominent, has aroused the people to the food safety strong concern. The food traceability system is considered to be an effective way to eliminate information asymmetry by creating a reliable and continuous information flow across the supply chain, monitoring the food production process and direction, and identifying the source of the problem and implementing the recall through traceability. One of the main tools for fundamental prevention of food safety risks. This paper studies consumers' willingness to pay for traceability attributes. In view of the lack of large-scale food safety research data in China, this paper studies the traceability information and levels of food traceability, and designs a questionnaire by menu selection experiment. The results of the research are input as the data of the research and analysis. Because menu method data is the characteristic of discrete classification attribute data, the general method is difficult to apply. In order to solve the above problems, this paper studies consumers' willingness to pay for pork traceability attributes by using the latent category model of economics and k-modes clustering algorithm, which is suitable for classification attribute data, so as to analyze consumers' group preference. In order to expand the consumer demand for traceability food, better promotion of food traceability system, food safety. The main work of this paper is as follows: (1) through the literature research and field investigation of domestic and foreign food traceability systems and willingness to pay, the design of pork supply chain system traceability information should include breeding, Slaughtering processing and distribution sales and government certification four attributes. The above four traceability attributes are set up with different price level design menu to choose the experimental questionnaire. The D-efficiency test shows that the questionnaire design is excellent. The data were collected in Wuxi City, Jiangsu Province, and the results of statistical analysis showed that the results were good. (2) according to the consumer utility theory, a latent class model is established for consumers' willingness to pay retroactively, taking the consumer's category as the potential variable and the consumer's choice as the explicit variable. The method of menu selection experiment was used to analyze consumer behavior. The results show that consumers' demand for food traceability attribute presents different preferences, and generally belongs to the low-level traceability consumer group. (3) according to the characteristic that menu selection experiment data is discrete classification attribute data, the k-modes algorithm suitable for clustering analysis is found through the research of clustering correlation algorithm. In view of the complex clustering process and the low classification accuracy of k-modes algorithm, the clustering process is simplified by selecting the initial clustering center by combining the density and distance factors, combined with the latest research progress, and the k-modes clustering is improved synthetically by combining the two factors of density and distance. The modes, of k-modes clustering algorithm is replaced by the pattern considering all attribute values of traceability attributes, so as to improve the accuracy of clustering classification. (4) the cluster analysis model is established according to the questionnaire, and the improved k-modes clustering method is applied to the analysis of the results of the menu selection experiment. The CU and objective functions are used to select the appropriate number of cluster categories. The results show that consumers can be divided into several groups with different preference for traceability security information, and these groups have different ability to pay. Different groups of pork can be provided with different traceability properties to expand consumers' demand for traceability food and improve food safety.
【學位授予單位】：江南大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TS201.6;TP311.13
本文編號：2438688
[Abstract]:In recent years, the domestic food safety quality problem is very prominent, has aroused the people to the food safety strong concern. The food traceability system is considered to be an effective way to eliminate information asymmetry by creating a reliable and continuous information flow across the supply chain, monitoring the food production process and direction, and identifying the source of the problem and implementing the recall through traceability. One of the main tools for fundamental prevention of food safety risks. This paper studies consumers' willingness to pay for traceability attributes. In view of the lack of large-scale food safety research data in China, this paper studies the traceability information and levels of food traceability, and designs a questionnaire by menu selection experiment. The results of the research are input as the data of the research and analysis. Because menu method data is the characteristic of discrete classification attribute data, the general method is difficult to apply. In order to solve the above problems, this paper studies consumers' willingness to pay for pork traceability attributes by using the latent category model of economics and k-modes clustering algorithm, which is suitable for classification attribute data, so as to analyze consumers' group preference. In order to expand the consumer demand for traceability food, better promotion of food traceability system, food safety. The main work of this paper is as follows: (1) through the literature research and field investigation of domestic and foreign food traceability systems and willingness to pay, the design of pork supply chain system traceability information should include breeding, Slaughtering processing and distribution sales and government certification four attributes. The above four traceability attributes are set up with different price level design menu to choose the experimental questionnaire. The D-efficiency test shows that the questionnaire design is excellent. The data were collected in Wuxi City, Jiangsu Province, and the results of statistical analysis showed that the results were good. (2) according to the consumer utility theory, a latent class model is established for consumers' willingness to pay retroactively, taking the consumer's category as the potential variable and the consumer's choice as the explicit variable. The method of menu selection experiment was used to analyze consumer behavior. The results show that consumers' demand for food traceability attribute presents different preferences, and generally belongs to the low-level traceability consumer group. (3) according to the characteristic that menu selection experiment data is discrete classification attribute data, the k-modes algorithm suitable for clustering analysis is found through the research of clustering correlation algorithm. In view of the complex clustering process and the low classification accuracy of k-modes algorithm, the clustering process is simplified by selecting the initial clustering center by combining the density and distance factors, combined with the latest research progress, and the k-modes clustering is improved synthetically by combining the two factors of density and distance. The modes, of k-modes clustering algorithm is replaced by the pattern considering all attribute values of traceability attributes, so as to improve the accuracy of clustering classification. (4) the cluster analysis model is established according to the questionnaire, and the improved k-modes clustering method is applied to the analysis of the results of the menu selection experiment. The CU and objective functions are used to select the appropriate number of cluster categories. The results show that consumers can be divided into several groups with different preference for traceability security information, and these groups have different ability to pay. Different groups of pork can be provided with different traceability properties to expand consumers' demand for traceability food and improve food safety.
【學位授予單位】：江南大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TS201.6;TP311.13
【參考文獻】
相關期刊論文 前10條
1 吳林海;高寧;朱淀;;不同消費群體對不同層次可追溯食品的需求研究[J];財貿(mào)研究;2013年05期
2 梁吉業(yè);白亮;曹付元;;基于新的距離度量的K-Modes聚類算法[J];計算機研究與發(fā)展;2010年10期
3 劉軍弟;王凱;韓紀琴;;消費者對有機豬肉的認知水平及其消費行為調(diào)研——基于上海與南京的調(diào)查數(shù)據(jù)[J];現(xiàn)代經(jīng)濟探討;2009年04期
4 李仁侃;葉東毅;;粗糙K-Modes聚類算法[J];計算機應用;2011年01期
5 梁穎;盧海燕;劉賢金;;食品安全認證現(xiàn)狀及其在我國的應用分析[J];江蘇農(nóng)業(yè)科學;2012年06期
6 陳默;王曉莉;吳林海;;R&D投入能力、企業(yè)特征、政府作用與企業(yè)低碳生產(chǎn)意愿研究[J];科技進步與對策;2010年22期
7 周莉;劉明春;;食品可追溯體系研究現(xiàn)狀[J];糧食與油脂;2008年07期
8 馬從國;趙德安;劉葉飛;倪軍;張玉峰;;豬肉工廠化生產(chǎn)的全程監(jiān)控與可溯源系統(tǒng)研制[J];農(nóng)業(yè)工程學報;2008年09期
9 張可;柴毅;翁道磊;翟茹玲;;豬肉生產(chǎn)加工信息追溯系統(tǒng)的分析和設計[J];農(nóng)業(yè)工程學報;2010年04期
10 孫吉貴;劉杰;趙連宇;;聚類算法研究[J];軟件學報;2008年01期
本文編號：2438688
本文鏈接：http://www.sikaile.net/guanlilunwen/gongyinglianguanli/2438688.html
最近更新
教材專著
