26計算機專業(yè)相關有關外文文獻翻譯成品基于消費者行為建模的網頁內容推薦系統(tǒng)

1、<p>　　外文標題：Web Content Recommender System based on Consumer Behavior Modeling</p><p>　　外文作者：ACM Fong ， B Zhou ， SC Hui ， GY Hong ， TA Do</p><p>　　文獻出處:《IEEE Transactions on Consumer Electro

2、nics》 , 2011 , 57 (2) :962-969</p><p>　　英文1595單詞， 8495字符，中文2689 漢字。</p><p>　　此文檔是外文翻譯成品，無需調整復雜的格式哦！下載之后直接可用，方便快捷！只需二十多元。</p><p>　　Web Content Recommender System based on Consumer Be

3、havior Modeling</p><p>　　A. C. M. Fong, Senior Member, IEEE, Baoyao Zhou, S. C. Hui, Senior Member, IEEE,Guan Y. Hong, and The Anh Do</p><p>　　Abstract — Web surfing has become a popular activit

4、y for many consumers who not only make purchases online, but also seek relevant information on products and services before they commit to buy. The authors propose a web recommender that models user habits and behaviors

5、by constructing a knowledge base using temporal web access patterns as input. Fuzzy logic is applied to represent real-life temporal concepts and requested resources of periodic pattern-based web access activities. The f

6、uzzy repr</p><p>　　Index Terms — Consumer behavior modeling, personalization, web content recommender, consumer internet application.</p><p>　　I. INTRODUCTION</p><p>　　The web has b

7、ecome an increasingly popular medium for consumer to exchange or find ideas, opinions, experiences on products and services. Many consumers go further than online information sharing and actually perform purchases on the

8、 web. Increasing availability and popularity of portable web- enabled handheld devices looks set to fuel further growth in the volume of consumer web traffic.</p><p>　　In the traditional web information diss

9、emination model, consumers are familiar with pulling content from the web via mechanisms such as search engines, or simply by typing the URI/URL if they happen to know it. Researchers have long tried to make web search m

10、ore efficient on handheld mobile devices e.g. [1], which generally have limited processing power and screen size compared to personal computers.</p><p>　　II.Related Work</p><p>　　This section pr

11、esents a survey of work pertinent to personalized web content recommendation and consumers’ web access pattern mining based on periodicity information. This section therefore serves to lay the foundation for further disc

12、ussion of the proposed approach in Section III.</p><p>　　A.Personalized Web Content Recommendation</p><p>　　Personalized web content recommendation aims at minimizing ambiguity and unwanted info

13、rmation that is presented to the consumer, thereby reducing the effect of information overload that is often encountered by web surfers. According to a survey presented in [10], traditional web content recommender system

14、s could be classified into Content-based, Collaborative and Hybrid, which is combination of the two. However, these systems tend to rely heavily on user ratings. Non-intrusiveness has been ident</p><p>　　The

15、 fundamental requirement of an effective personalized web content recommendation system is to present the most relevant suggestions to the user in a timely manner. Thus, both context and temporal information is important

16、. In this regard, context information can be very effective in disambiguation. For example, a biologist who wants to read articles about “mouse” will likely be interested in the rodent. On the other hand, a consumer look

17、ing for computer accessories will likely be interested in</p><p>　　Currently systems tend to lack focus on temporal information. Timeliness in the recommendation of relevant resources is also important in ma

18、ny situations. For example, a user may have a tendency of reading financial news and traffic information between 9am and 10am on weekdays (or even more specifically on Mondays and Tuesdays), but may tend to read entertai

19、nment news and weather information during the same time period on Saturdays.</p><p>　　III.Consumer Behavior Knowledge Base</p><p>　　The process of constructing the user behavior knowledge base b

20、egins with semantically enhanced web usage logs as input. The raw log data are then preprocessed to remove unnecessary information and to enhance the quality of the data for subsequent processing. Finally, one can procee

21、d to construct the knowledge base using fuzzy relations as the basis. As mentioned in Section I, fuzzy representation can better describe real-life situations than a standard crisp binary representation.</p><p

22、>　　A.Semantically Enhanced Web Usage Logs</p><p>　　In the current Web environment, web usage logs (including the web server logs, browser logs and proxy logs) record access requests from users to one or m

23、ultiple websites as a sequence of requested URLs with timestamps. In this research, the focus is on web server logs. However, the URLs recorded in web usage logs contain little semantic information about the web content

24、accessed by users. This makes it difficult to be used for understanding users' actual access behaviors, interests and intention</p><p>　　To overcome this problem, the periodic association access patterns

25、 should be filtered to retain only those interesting patterns, which are more important for describing the periodic web access behavior of the user. This is easily done by pruning the periodic association access patterns

26、 with low Sup and Conf values. In the experiments, it was found that the thresholds for these value should be about 0.1 - 0.15, depending on the actual application.</p><p>　　VI.Conclusion</p><p>

27、;　　The authors have proposed an approach for constructing a user behavior knowledge base, which uses fuzzy logic to represent real-life temporal concepts and meaningful resources for periodic pattern-based web access act

28、ivities. Bother objective and subjective tests have been conducted. The experimental results have shown that the proposed approach can achieve effective periodic web personalization.</p><p>　　By performing c

29、ompute-intensive behavior analysis, modeling and knowledge base construction in advance, future application may include real-time recommendation on portable web-enabled devices that are becoming increasingly popular amon

30、g consumers of electronics products, e.g. cellular phones that have limited processing power relative to a personal computer would be well suited to this kind of asymmetric approach. In the future, consumers who choose t

31、o use this system on their web- enabled mobil</p><p>　　References</p><p>　　[1]W. Lee, S. Kang; S. Lim, M.-K. Shin and Y.-K. Kim, “Adaptive hierarchical surrogate for searching web with mobile de

32、vices”, IEEE Trans. Consumer Electron., vol. 53, no. 2, pp. 796 - 803, 2007.</p><p>　　[2]Y.B. Fernandez, J.J.P. Arias, M.L. Nores, A.G. Solla and M.R. Cabrer, “AVATAR: an improved solution for personalized T

33、V based on semantic inference,” IEEE Trans. Consumer Electron., vol. 52, no. 1, pp. 223 - 231, 2006.</p><p>　　[3]A. Martinez, J. Arias, A. Vilas, J. Garcia Duque, M. Lopez Nores, “What's on TV tonight? A

34、n efficient and effective personalized recommender system of TV programs,” IEEE Trans. Consumer Electron., vol. 55, no. 1, pp. 286 - 294, 2009.</p><p>　　[4]S. Lee, D. Lee and S. Lee, “Personalized DTV progra

35、m recommendation system under a cloud computing environment,” IEEE Trans. Consumer Electron., vol. 56, no. 2, pp. 1034 - 1042, 2010.</p><p>　　[5]H. Lee; S. Lee, H. Kim, H. Bahn, “Personalized recommendation

36、schemes for DTV channel selectors”, IEEE Trans. Consumer Electron., vol. 52, no. 3, pp. 1064 - 1068, 2006.</p><p>　　[6]H. Shin, M. Lee and E. Kim, “Personalized digital TV content recommendation with integra

37、tion of user behavior profiling and multimodal content rating,” IEEE Trans. Consumer Electron., vol. 55, no. 3, pp. 1417 - 1423, 2009.</p><p>　　[7]S. Park, J. Jeong, H. Jo, J. Lee and E. Seo, “Development of

38、 behavior- profilers for multimedia consumer electronics,” IEEE Trans. Consumer Electron., vol. 55, no. 4, pp. 1929 - 1935, 2009.</p><p>　　[8]J. Srivastava, R. Cooley, M. Deshpande, and P.-N. Tan, “Web usage

39、 mining: discovery and applications of usage patterns from web data”, ACM SIGKDD Explorations, 1(2), 2000, pp. 12-23.</p><p>　　[9]J. Pei, J. Han, B. Mortazavi-Asl, and H. Zhu, “Mining access patterns efficie

40、ntly from web logs”，Proc. 4th Pacific-Asia Conf. Knowledge Discovery and Data Mining, Kyoto, Japan, 2000, pp. 396-407.</p><p>　　[10]G. Adomavicius and A. Tuzhilin, “Toward the next generation of recommender

41、systems: A survey of the state-of-the-art and possible extensions”, IEEE Trans. Knowl & Data Eng, 17(6), 2005, pp. 734-749.</p><p>　　[11]B. Ozden, S. Ramaswamy, and A. Silberschatz, “Cyclic Association R

42、ules”, In ICDE ’98: Proc. 14th International Conference on Data Engineering, pp. 412421, Orlando, Florida, USA, 1998.</p><p>　　[12]S. Ramaswamy, S. Mahajan, and A. Silberschatz, “On the Discovery of interest

43、ing patterns in association rules”, In VLDB ’98: Proc. 24th International Conference on Very Large Data Bases, pp.368-379, New York, USA, Aug. 1998.</p><p>　　[13]Y. Li, P. Ning, X. S. Wang, and S. Jajodia, “

44、Discovering calendar-based temporal association rules”, In TIME ’01: Proc. 8th International Symposium on Temporal Representation and Reasoning, pp. 111-118, Civdale del Friuli, Italy, June 2001</p><p>　　[14

45、]Y. Li, P. Ning, X. S. Wang, and S. Jajodia, “Discovering calendar-based temporal association rules”, Data & Knowledge Engineering, 44(2), pp. 193-218, 2003.</p><p>　　基于消費者行為建模的網頁內容推薦系統(tǒng)</p><p&

46、gt;<b>　　摘要</b></p><p>　　對許多消費者而言，網上沖浪已經成為其普遍的活動，不僅可以在網上購物，而且還在購買之前還可尋求有關產品和服務的信息。本文中，作者提出了一種網絡推薦系統(tǒng)，它通過使用當時的網絡訪問模式作為輸入來構建知識數(shù)據庫以模擬用戶的習慣和行為。應用模糊邏輯來表示實際當時的概念并請求周期性基于模式的Web訪問活動的資源。模糊表示用于構建用戶網絡訪問習慣和行為的

47、知識數(shù)據庫，用于向用戶提供及時的個性化推薦。其所提出的方法適用于為消費者的便攜式設備提供建議，因為計算機密集型處理過程是離線的和事先進行的。隨著網絡消費移動設備的可用性和普及程度越來越高，人們相信未來的CE世界將越來越面向網絡。為了評估所提出方法的性能表現(xiàn)而進行的實驗已經顯示出非常令人滿意的結果。</p><p>　　索引詞 - 消費者行為建模，個性化，網頁內容推薦，消費者互聯(lián)網應用。</p>&l

48、t;p><b>　　1引言</b></p><p>　　對消費者而言，網絡日益成為其交流或尋找產品和服務想法、意見和經驗的的流行媒介。 許多消費者分享的信息比網上信息更實際，并實際上就在網上進行購買。 便攜式網絡手持設備的可用性和普及度的增加似乎將推動消費者網絡流量的進一步增長。</p><p>　　在傳統(tǒng)的Web信息傳播模式中，消費者熟悉通過諸如搜索引擎之類的

49、機制從網絡上獲取內容，或者只需輸入網址鏈接URI / URL（如果他們碰巧知道它）。 研究人員一直試圖使手持移動設備上的網絡以使搜索更高效，例如， [1]，與個人電腦相比，它的處理能力和屏幕尺寸一般都很有限。</p><p><b>　　2相關工作</b></p><p>　　本節(jié)介紹了基于周期性信息的個性化Web內容推薦和消費者Web訪問模式挖掘相關工作的調查。 因

50、此，本節(jié)為在第三節(jié)進一步討論擬議的方法打下基礎。</p><p><b>　　個性化網頁內容推薦</b></p><p>　　個性化網絡內容推薦旨在最大程度地減少呈現(xiàn)給消費者模糊和不必要的信息，從而降低網民經常遇到的信息過載的影響。 根據文獻[10]中的一項調查，傳統(tǒng)的網絡內容推薦系統(tǒng)可以分為基于內容的，協(xié)作的和混合的，這兩者相結合。 但是，這些系統(tǒng)往往嚴重依賴于用戶

51、評分。 無干擾性已被確定為后續(xù)網絡內容推薦信息收集過程中的重要屬性[10]。 網絡使用信息的挖掘，由系統(tǒng)在后臺執(zhí)行，并對用戶透明，因此它代表了這一研究領域的重要發(fā)展方向。</p><p>　　有效的個性化網頁內容推薦系統(tǒng)的基本要求是及時向用戶提供最相關的建議。因此，情境和時間信息都很重要。在這方面，背景信息在消除模糊信息方面非常有效。例如，想要閱讀關于“老鼠”的文章的生物學家可能會對嚙齒動物感興趣。另一方面，尋找

52、電腦配件的消費者可能對定點設備感興趣。前后的信息將有助于解決此類模糊的信息。</p><p>　　目前一些系統(tǒng)往往缺乏對時間信息的關注。相關資源建議的及時性在許多情況下也很重要。例如，用戶可能傾向于在平日上午9點至上午10點（甚至更具體地在星期一和星期二）閱讀金融新聞和交通信息，但可能傾向于在周六的相同時間段內閱讀娛樂新聞和天氣信息。</p><p>　　3消費者行為信息數(shù)據庫</p

53、><p>　　構建用戶行為信息數(shù)據庫的過程始于語義增強的Web使用日志作為輸入。然后對原始日志數(shù)據進行預處理以去除不必要的信息并提高后續(xù)處理的數(shù)據質量。最后，可以用模糊關系作為基礎來構建信息數(shù)據庫。正如第一部分提到的，模糊表示可以比標準的簡潔二進制表示能更好地描述真實情況。</p><p>　　A.語義增強的Web使用情況日志</p><p>　　在當前的Web環(huán)境中，

54、Web使用情況日志（包括Web服務器日志，瀏覽器日志和代理日志）將用戶訪問一個或多個網站的訪問請求記錄為具有時間戳的所請求的URL序列。在這項研究中，重點是Web服務器日志。但是，Web使用日志中記錄的URL包含的關于用戶訪問的Web內容的語義信息很少。這對理解用戶的實際訪問行為、興趣和意圖變得很難。因此，需要某種形式的語義增強功能才能使Web日志數(shù)據真正有用。</p><p>　　B.信息數(shù)據庫的建設</

55、p><p>　　上述語義增強的Web使用日志現(xiàn)在可被用作消費者行為信息數(shù)據庫構建過程的輸入。圖1總結了知識庫構建過程中進一步處理的四個關鍵步驟。針對傳統(tǒng)（非語義增強）Web服務器日志中[19]中討論的類似預處理任務、數(shù)據清理、用戶標識和會話標識進行了修改。特別是，在輸入的Web使用情況日志中，并非所有記錄對于挖掘目的都是必需的或有用的。根據所選資源屬性集合，根據“語義標注”字段中的主題，清除不必要的條目非常重要。當輸

56、入日志中的條目涉及至少一個其他所有記錄中的資源屬性時，將從Web使用日志中丟棄，其中包括圖片文件、腳本和其他無效文檔。此外，包含不成功請求條目的網頁使用日志也被丟棄。</p><p>　　為了克服這個問題，應該對周期性關聯(lián)訪問模式進行過濾，以僅保留那些有趣的模式，這對于描述用戶的周期性Web訪問行為更為重要。 這很容易通過修剪具有低Sup和Conf值的周期性關聯(lián)訪問模式來完成。 在實驗中，發(fā)現(xiàn)這些值的閾值范圍應為

57、約0.1-0.15，這取決于其實際應用。</p><p><b>　　4 結論</b></p><p>　　在本文中，作者提出了一種構建用戶行為知識數(shù)據庫的方法，該方法使用模糊邏輯來表示基于周期性模式的Web訪問活動的真實時間概念及有意義的資源。并進行了客觀和主觀性的測試。實驗結果表明，該方法可以實現(xiàn)網絡個性化推薦有效性和周期性。</p><p&g

58、t;　　通過預先進行計算密集型行為的分析、建模和知識庫的構建，未來的應用可以包括對在電子產品的消費者中變得越來越流行的便攜式網絡使能設備進行實時推薦，例如，具有相對于個人計算機的處理能力有限的蜂窩電話而言，它將非常適合這種非對稱方法。未來，選擇在其網絡移動設備上使用此系統(tǒng)的消費者將能夠花費更少的時間和精力來搜索他們想要的內容，并可以在適當?shù)臅r候獲得更多可能需要推薦給他們的內容。</p><p><b>

59、　　參考文獻</b></p><p>　　[1]W. Lee, S. Kang; S. Lim, M.-K. Shin and Y.-K. Kim, “Adaptive hierarchical surrogate for searching web with mobile devices”, IEEE Trans. Consumer Electron., vol. 53, no. 2, pp. 79

60、6 - 803, 2007.</p><p>　　[2]Y.B. Fernandez, J.J.P. Arias, M.L. Nores, A.G. Solla and M.R. Cabrer, “AVATAR: an improved solution for personalized TV based on semantic inference,” IEEE Trans. Consumer Electron.

61、, vol. 52, no. 1, pp. 223 - 231, 2006.</p><p>　　[3]A. Martinez, J. Arias, A. Vilas, J. Garcia Duque, M. Lopez Nores, “What's on TV tonight? An efficient and effective personalized recommender system of T

62、V programs,” IEEE Trans. Consumer Electron., vol. 55, no. 1, pp. 286 - 294, 2009.</p><p>　　[4]S. Lee, D. Lee and S. Lee, “Personalized DTV program recommendation system under a cloud computing environment,”

63、IEEE Trans. Consumer Electron., vol. 56, no. 2, pp. 1034 - 1042, 2010.</p><p>　　[5]H. Lee; S. Lee, H. Kim, H. Bahn, “Personalized recommendation schemes for DTV channel selectors”, IEEE Trans. Consumer Elect

64、ron., vol. 52, no. 3, pp. 1064 - 1068, 2006.</p><p>　　[6]H. Shin, M. Lee and E. Kim, “Personalized digital TV content recommendation with integration of user behavior profiling and multimodal content rating,

65、” IEEE Trans. Consumer Electron., vol. 55, no. 3, pp. 1417 - 1423, 2009.</p><p>　　[7]S. Park, J. Jeong, H. Jo, J. Lee and E. Seo, “Development of behavior- profilers for multimedia consumer electronics,” IEE

66、E Trans. Consumer Electron., vol. 55, no. 4, pp. 1929 - 1935, 2009.</p><p>　　[8]J. Srivastava, R. Cooley, M. Deshpande, and P.-N. Tan, “Web usage mining: discovery and applications of usage patterns from web

67、 data”, ACM SIGKDD Explorations, 1(2), 2000, pp. 12-23.</p><p>　　[9]J. Pei, J. Han, B. Mortazavi-Asl, and H. Zhu, “Mining access patterns efficiently from web logs”，Proc. 4th Pacific-Asia Conf. Knowledge Dis

68、covery and Data Mining, Kyoto, Japan, 2000, pp. 396-407.</p><p>　　[10]G. Adomavicius and A. Tuzhilin, “Toward the next generation of recommender systems: A survey of the state-of-the-art and possible extensi

69、ons”, IEEE Trans. Knowl & Data Eng, 17(6), 2005, pp. 734-749.</p><p>　　[11]B. Ozden, S. Ramaswamy, and A. Silberschatz, “Cyclic Association Rules”, In ICDE ’98: Proc. 14th International Conference on Dat

70、a Engineering, pp. 412421, Orlando, Florida, USA, 1998.</p><p>　　[12]S. Ramaswamy, S. Mahajan, and A. Silberschatz, “On the Discovery of interesting patterns in association rules”, In VLDB ’98: Proc. 24th In

71、ternational Conference on Very Large Data Bases, pp.368-379, New York, USA, Aug. 1998.</p><p>　　[13]Y. Li, P. Ning, X. S. Wang, and S. Jajodia, “Discovering calendar-based temporal association rules”, In TIM

72、E ’01: Proc. 8th International Symposium on Temporal Representation and Reasoning, pp. 111-118, Civdale del Friuli, Italy, June 2001</p><p>　　[14]Y. Li, P. Ning, X. S. Wang, and S. Jajodia, “Discovering cale