【摘要】：移動(dòng)電子商務(wù)是在無線平臺(tái)上實(shí)現(xiàn)的電子商務(wù)。近年來移動(dòng)電子商務(wù)由于其方便快捷的支付方式以及隨時(shí)隨地提供服務(wù)的優(yōu)勢(shì)，得到了迅猛發(fā)展；同時(shí)隨著4G網(wǎng)絡(luò)和智能手機(jī)的普及，移動(dòng)電子商務(wù)具有極大發(fā)展?jié)摿ΑＯ啾扔谶\(yùn)行在PC端的傳統(tǒng)電子商務(wù)，移動(dòng)平臺(tái)在屏幕上所能展示的商品信息相對(duì)有限，如何讓用戶迅速找到自己感興趣和需要的商品，避免信息超載，而不是迷失在大量商品信息中，已成為移動(dòng)電子商務(wù)發(fā)展的一個(gè)亟待解決的問題。推薦系統(tǒng)是解決信息超載問題的一個(gè)有效方法，它通過分析用戶的興趣偏好、個(gè)性化需求等，使用關(guān)聯(lián)規(guī)則、協(xié)同過濾等推薦技術(shù)向用戶推薦個(gè)性化信息。然而在實(shí)際應(yīng)用中，已有的電子商務(wù)推薦系統(tǒng)仍存在著一些問題，如稀疏問題，可擴(kuò)展，模型過擬合等問題，導(dǎo)致推薦效率較低，推薦質(zhì)量不高，不能夠滿足用戶的個(gè)性化需求，因此，對(duì)于移動(dòng)電子商務(wù)推薦系統(tǒng)和推薦技術(shù)的研究具有比較大的實(shí)用價(jià)值。 本文以實(shí)際的移動(dòng)電子商務(wù)系統(tǒng)為應(yīng)用背景，通過分析移動(dòng)電子商務(wù)推薦系統(tǒng)的特征和當(dāng)前推薦系統(tǒng)在移動(dòng)電子商務(wù)和海量數(shù)據(jù)環(huán)境下存在的實(shí)時(shí)性和推薦效率不高等問題，對(duì)推薦系統(tǒng)的體系結(jié)構(gòu)、功能模塊以及工作流程進(jìn)行了學(xué)習(xí)研究，設(shè)計(jì)了基于關(guān)聯(lián)規(guī)則的移動(dòng)電子商務(wù)推薦系統(tǒng)模型，該模型將推薦過程分為離線處理和在線推薦兩大部分，離線處理又分為數(shù)據(jù)預(yù)處理和關(guān)聯(lián)規(guī)則挖掘兩個(gè)子模塊。數(shù)據(jù)預(yù)處理通過數(shù)據(jù)庫觸發(fā)器和存儲(chǔ)過程來實(shí)現(xiàn)數(shù)據(jù)的選擇清理和格式的轉(zhuǎn)換，，關(guān)聯(lián)規(guī)則挖掘部分采用FP-growth算法實(shí)現(xiàn)頻繁模式的挖掘，生成并導(dǎo)入關(guān)聯(lián)規(guī)則庫。在線推薦模塊根據(jù)采集的用戶信息與生成的規(guī)則庫產(chǎn)生準(zhǔn)確、實(shí)時(shí)的個(gè)性化推薦結(jié)果。該模型在推薦效率、推薦質(zhì)量上進(jìn)行了有益研究，推薦模型在移動(dòng)電子商務(wù)系統(tǒng)中的應(yīng)用有效地提高了推薦效率和實(shí)時(shí)性，能夠很好地為用戶推薦符合其興趣偏好和需求的商品，從而提高商品銷量和用戶忠誠度。
[Abstract]:Mobile e-commerce is an e-commerce implemented on wireless platform. In recent years, mobile electronic commerce has developed rapidly because of its convenient and quick payment method and the advantage of providing services anytime and anywhere. With the popularity of 4G network and smart phone, mobile electronic commerce has great development potential. Compared with the traditional e-commerce running on the PC, the mobile platform can display relatively limited information on the screen, how to quickly find the products that users are interested in and need, and avoid information overload. Instead of getting lost in a lot of commodity information, it has become an urgent problem in the development of mobile e-commerce. Recommendation system is an effective method to solve the problem of information overload. It recommends personalized information to users by analyzing users' interest preferences, personalized requirements, using association rules, collaborative filtering and other recommendation techniques. However, in the practical application, there are still some problems in the existing E-commerce recommendation system, such as sparse problem, extensibility, model over-fitting and so on, which lead to low recommendation efficiency and low recommendation quality. Therefore, the research on mobile e-commerce recommendation system and recommendation technology is of great practical value. This paper takes the actual mobile electronic commerce system as the application background, through the analysis of the characteristics of the mobile electronic commerce recommendation system and the current recommendation system in the mobile electronic commerce and mass data environment of real-time and recommendation efficiency is not high, and so on. The architecture, function modules and workflow of the recommendation system are studied, and a mobile e-commerce recommendation system model based on association rules is designed. The model divides the recommendation process into two parts: offline processing and online recommendation. Off-line processing is divided into two sub-modules: data preprocessing and association rule mining. Data preprocessing realizes data selection and format conversion by database triggers and stored procedures, and association rules mining uses FP-growth algorithm to mine frequent patterns and generate and import association rules database. The online recommendation module generates accurate and real-time personalized recommendation results according to the collected user information and the generated rule base. This model has carried on the beneficial research in the recommendation efficiency, the recommendation quality, the recommendation model in the mobile electronic commerce system application has effectively improved the recommendation efficiency and the real-time performance. It can recommend the products according to their interests, preferences and needs to improve the sales volume and customer loyalty.
【學(xué)位授予單位】：蘭州交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TP391.3

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 鄭斌;李涓子;;一種基于FP-Growth的改進(jìn)算法[J];平頂山工學(xué)院學(xué)報(bào);2008年04期

2 吳吉義;林志潔;龔祥國;;基于協(xié)同過濾的移動(dòng)電子商務(wù)個(gè)性化推薦系統(tǒng)若干研究[J];電子技術(shù)應(yīng)用;2007年01期

3 索琪,盧濤;基于關(guān)聯(lián)規(guī)則的電子商務(wù)推薦系統(tǒng)研究[J];哈爾濱師范大學(xué)自然科學(xué)學(xué)報(bào);2005年02期

4 李忠俊;周啟海;帥青紅;;一種基于內(nèi)容和協(xié)同過濾同構(gòu)化整合的推薦系統(tǒng)模型[J];計(jì)算機(jī)科學(xué);2009年12期

5 楊引霞,謝康林,朱揚(yáng)勇,左子葉;電子商務(wù)網(wǎng)站推薦系統(tǒng)中關(guān)聯(lián)規(guī)則推薦模型的實(shí)現(xiàn)[J];計(jì)算機(jī)工程;2004年19期

6 閆鶯,王大玲,于戈;支持個(gè)性化推薦的Web頁面關(guān)聯(lián)規(guī)則挖掘算法[J];計(jì)算機(jī)工程;2005年01期

7 馬建慶;鐘亦平;張世永;;基于興趣度的關(guān)聯(lián)規(guī)則挖掘算法[J];計(jì)算機(jī)工程;2006年17期

8 丁振國,陳靜;基于關(guān)聯(lián)規(guī)則的個(gè)性化推薦系統(tǒng)[J];計(jì)算機(jī)集成制造系統(tǒng)-CIMS;2003年10期

9 余力,劉魯,李雪峰;用戶多興趣下的個(gè)性化推薦算法研究[J];計(jì)算機(jī)集成制造系統(tǒng);2004年12期

10 姚忠;吳躍;常娜;;集成項(xiàng)目類別與語境信息的協(xié)同過濾推薦算法[J];計(jì)算機(jī)集成制造系統(tǒng);2008年07期

本文編號(hào)：2257060