發(fā)布時(shí)間：2018-06-19 18:35

  本文選題：商戶評分?jǐn)?shù)據(jù) + 關(guān)聯(lián)分析��； 參考：《鄭州大學(xué)》2017年碩士論文

【摘要】：當(dāng)今社會,人們在出門選擇商家或者購買商品之前,往往會先查看電商網(wǎng)站上的商戶評分信息,從而判斷這家商戶的服務(wù)或產(chǎn)品質(zhì)量等情況,以便做出消費(fèi)決策。人們經(jīng)常會使用網(wǎng)站上的商戶評分信息進(jìn)行消費(fèi)參考,而很少有人研究這些商戶評分信息本身。對商戶的評分?jǐn)?shù)據(jù)進(jìn)行合理的分析研究,從而發(fā)現(xiàn)數(shù)據(jù)之間的關(guān)聯(lián)關(guān)系,有助于用戶判斷商戶評分?jǐn)?shù)據(jù)的可信度,做出合理的消費(fèi)選擇。在本文研究中,筆者從關(guān)聯(lián)分析的角度出發(fā),分析這些電商網(wǎng)站上的商戶評分?jǐn)?shù)據(jù)。首先介紹商戶的評分?jǐn)?shù)據(jù)、關(guān)聯(lián)分析、相關(guān)分析、回歸分析的基礎(chǔ)概念與方法,然后利用相關(guān)分析和回歸分析的方法,設(shè)計(jì)一套適用于電商網(wǎng)站商戶評分?jǐn)?shù)據(jù)的關(guān)聯(lián)分析方法,該方法分為數(shù)據(jù)采集、數(shù)據(jù)清洗和關(guān)聯(lián)分析三個(gè)步驟,其中關(guān)聯(lián)分析部分又分為三個(gè)方面,分別是商戶評分?jǐn)?shù)據(jù)的相關(guān)分析、線性回歸分析和虛擬回歸分析,利用相關(guān)分析找出商戶各項(xiàng)評分之間或商戶評分與人均消費(fèi)之間的關(guān)聯(lián),利用線性回歸方法分析商戶總評分與各項(xiàng)評分之間的關(guān)系,利用虛擬回歸分析探索不同類別對于商戶評分的影響,找出商戶評分?jǐn)?shù)據(jù)之間的關(guān)聯(lián)關(guān)系。為了驗(yàn)證這套商戶評分?jǐn)?shù)據(jù)的關(guān)聯(lián)分析方法,筆者采集5個(gè)地市的3700條餐飲商戶評分?jǐn)?shù)據(jù),對這套關(guān)聯(lián)分析方法進(jìn)行實(shí)證研究。利用相關(guān)分析發(fā)現(xiàn)商戶各項(xiàng)評分之間、人均消費(fèi)和商戶各項(xiàng)評分之間的關(guān)聯(lián)關(guān)系,找出用戶的打分習(xí)慣,分析人均消費(fèi)與商戶各方面服務(wù)是否存在關(guān)聯(lián);利用線性回歸分析計(jì)算出商戶總評分與各方面評分之間的關(guān)聯(lián)關(guān)系,判斷商戶總評分?jǐn)?shù)據(jù)的可信度,了解哪些因素更能影響商戶的總評分;利用虛擬回歸分析找出地域?qū)θ藗兿埠貌讼档挠绊�。查看�?shù)據(jù)的處理效果,驗(yàn)證這套商戶評分?jǐn)?shù)據(jù)的關(guān)聯(lián)分析方法。對于消費(fèi)者來說,可以驗(yàn)證日常生活中使用的商戶評分?jǐn)?shù)據(jù)是否可信,增加人們對這些評分?jǐn)?shù)據(jù)的了解;對于商戶來說,可以幫助他們找到消費(fèi)者的喜好,助其發(fā)展經(jīng)營。
[Abstract]:In today's society, people often check the score information on e-commerce websites before going out to choose merchants or buy goods, so as to judge the service or product quality of the merchants in order to make consumer decisions. People often use the business rating information on the website for consumer reference, but few people study the information itself. This paper makes a reasonable analysis and research on the score data of the merchant, and finds out the correlation between the data, which is helpful for the user to judge the credibility of the scoring data and make a reasonable consumption choice. In this study, the author analyzes the score data of these e-commerce websites from the perspective of correlation analysis. This paper first introduces the basic concepts and methods of scoring data, association analysis, correlation analysis and regression analysis of merchants, and then designs a set of association analysis methods suitable for the score data of merchants on e-commerce websites by using the methods of correlation analysis and regression analysis. The method is divided into three steps: data acquisition, data cleaning and association analysis. Correlation analysis was used to find out the relationship between the score of the merchant and the consumption per capita, and the linear regression method was used to analyze the relationship between the total score of the merchant and the score. Virtual regression analysis is used to explore the influence of different categories on the score of merchants and to find out the correlation between the data. In order to verify the correlation analysis method, the author collected 3700 restaurant rating data from 5 prefectures, and made an empirical study on the correlation analysis method. By using the correlation analysis, we find out the relationship between the scores, the consumption per capita and the scores, find out the scoring habits of the users, and analyze whether there is a correlation between the consumption per capita and the various aspects of services. The linear regression analysis is used to calculate the correlation between the total score of the merchant and the score of all aspects, to judge the reliability of the data of the total score of the merchant, and to find out which factors can affect the total score of the merchant. Virtual regression analysis was used to find out the influence of region on people's favorite cuisine. Check the effect of data processing, verify this set of business score data association analysis method. For consumers, it can verify the credibility of the business score data used in daily life, increase people's understanding of these scoring data; for businesses, it can help them to find consumer preferences and help them to develop business.
【學(xué)位授予單位】：鄭州大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：F724.6;F274

【相似文獻(xiàn)】

相關(guān)期刊論文 前1條

1 倪萌;蘇新義;;跳水比賽中評分?jǐn)?shù)據(jù)的統(tǒng)計(jì)研究[J];新西部(下半月);2009年04期

相關(guān)會議論文 前3條

1 羅芬;丁樹良;;用修正的MDIE估計(jì)IRT中未知參數(shù)[A];全國教育與心理統(tǒng)計(jì)測量學(xué)術(shù)年會論文摘要集[C];2006年

2 李良萍;王運(yùn)良;;跳水新規(guī)則單人項(xiàng)目計(jì)分方法的合理性探討——由第15屆世界杯跳水賽評分?jǐn)?shù)據(jù)引發(fā)的思考[A];第八屆全國體育科學(xué)大會論文摘要匯編（二）[C];2007年

3 李培娟;安永福;王繼華;劉松柏;;奶牛9分制體型線性評分?jǐn)?shù)據(jù)的正態(tài)性分析[A];中國牛業(yè)健康發(fā)展與科技創(chuàng)新——中國畜牧獸醫(yī)學(xué)會第七屆養(yǎng)牛學(xué)分會2009年學(xué)術(shù)研討會論文集[C];2009年

相關(guān)碩士學(xué)位論文 前2條

1 簡丹;商戶評分?jǐn)?shù)據(jù)的關(guān)聯(lián)分析方法與實(shí)證研究[D];鄭州大學(xué);2017年

2 唐光杰;科技評審中評分偏差模型的研究與應(yīng)用[D];中南大學(xué);2012年

，

本文編號：2040870