【摘要】： 隨著電子商務(wù)和網(wǎng)絡(luò)技術(shù)的迅猛發(fā)展,信息和商品的數(shù)量急劇增加,在此背景下推薦系統(tǒng)和個(gè)性化推薦技術(shù)應(yīng)運(yùn)而生。推薦系統(tǒng)能夠幫助客戶尋找其感興趣的商品,具有鞏固老客戶,吸引新客戶,提高用戶的滿意度的作用,從而提高銷售量。但是,商品數(shù)量和品種的增多的同時(shí),也出現(xiàn)了有些具有價(jià)值的商品不易被用戶發(fā)現(xiàn)而滯銷的問題,以至物資浪費(fèi),利益損失。挖掘滯銷商品的價(jià)值,不僅能夠降低企業(yè)的損失,還能憑借低成本創(chuàng)造更大的利益。因此,滯銷商品的推薦技術(shù)研究在電子商務(wù)中具有重要意義。本文首先介紹了幾種常用的個(gè)性化推薦技術(shù)及方法。重點(diǎn)介紹了基于用戶的協(xié)同過濾推薦技術(shù)的基本思想和相似鄰居的計(jì)算,以及關(guān)聯(lián)規(guī)則的應(yīng)用。 其次,通過分析歷史銷售數(shù)據(jù)來預(yù)測(cè)和挖掘滯銷商品及與之密切相關(guān)的中間商品,提出一種借助中間商品激勵(lì)用戶購(gòu)買滯銷商品的滯銷商品推薦模型,并提出按照時(shí)間權(quán)重的中間商品興趣度進(jìn)行項(xiàng)目的個(gè)性化推薦。此推薦模型能夠有效預(yù)測(cè)滯銷商品,并根據(jù)客戶特征進(jìn)行個(gè)性化推薦,為滯銷商品的推薦提供一種行之有效的方法。 此外,在此模型基礎(chǔ)上,提出一種利潤(rùn)最大化的滯銷商品推薦算法,其基本思想是通過商品的關(guān)聯(lián)性確定最大利潤(rùn)推薦項(xiàng)集,再根據(jù)利潤(rùn)函數(shù)為每個(gè)用戶推薦利潤(rùn)最高的項(xiàng)集,其中推薦項(xiàng)集由中間商品和滯銷商品組成。 最后,本文通過實(shí)驗(yàn)驗(yàn)證了滯銷商品預(yù)測(cè)方法,基于利潤(rùn)最大化的滯銷商品推薦算法的可行性。
[Abstract]:With the rapid development of electronic commerce and network technology, the quantity of information and commodity increases rapidly. Under this background, recommendation system and personalized recommendation technology emerge as the times require. Recommendation system can help customers to find the goods they are interested in. It can consolidate the old customers, attract new customers, improve customer satisfaction, and thus increase the sales volume. However, with the increase of the quantity and variety of goods, there are also some problems of unsalable goods which are not easy to be discovered by users, even material waste and loss of profits. Excavating the value of unsalable goods can not only reduce the loss of enterprises, but also create greater benefits by means of low cost. Therefore, the recommendation technology of unsalable goods is of great significance in e-commerce. This paper first introduces several commonly used personalized recommendation techniques and methods. The basic idea of user-based collaborative filtering recommendation technology, the computation of similar neighbors, and the application of association rules are introduced. Secondly, by analyzing the historical sales data to predict and excavate the unsalable goods and the intermediate commodities closely related to them, a recommendation model of unsalable commodities is proposed to encourage customers to buy unsalable goods with the help of intermediate commodities. And put forward according to the time weight of the intermediate commodity interest degree to carry on the item personalized recommendation. This recommendation model can effectively predict unsalable goods and make personalized recommendation according to customer characteristics, which provides an effective method for the recommendation of unsalable goods. In addition, on the basis of this model, a recommendation algorithm for unsalable goods with maximum profit is proposed. The basic idea is to determine the maximum profit recommendation item set through the correlation of goods, and then to recommend the highest profit set for each user according to the profit function. The recommended item set consists of intermediate commodities and unsalable commodities. Finally, this paper verifies the feasibility of unsalable goods prediction method and unsalable goods recommendation algorithm based on profit maximization.
【學(xué)位授予單位】：沈陽(yáng)航空工業(yè)學(xué)院
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2010
【分類號(hào)】：F713.36

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 趙亮,胡乃靜,張守志;個(gè)性化推薦算法設(shè)計(jì)[J];計(jì)算機(jī)研究與發(fā)展;2002年08期

2 周軍鋒,湯顯,郭景峰;一種優(yōu)化的協(xié)同過濾推薦算法[J];計(jì)算機(jī)研究與發(fā)展;2004年10期

3 馬海兵,張成洪,張錦,胡運(yùn)發(fā);基于IS~±樹模型的頻繁模式挖掘[J];計(jì)算機(jī)研究與發(fā)展;2005年04期

4 高瀅;齊紅;劉杰;劉大有;;結(jié)合似然關(guān)系模型和用戶等級(jí)的協(xié)同過濾推薦算法[J];計(jì)算機(jī)研究與發(fā)展;2008年09期

5 李峰;李軍懷;王瑞林;張t，

本文編號(hào)：2404938