本文選題：移動(dòng)用戶(hù)上網(wǎng)行為 + 4G網(wǎng)絡(luò)　； 參考：《北京交通大學(xué)》2014年碩士論文

【摘要】：隨著工信部4G牌照的發(fā)放,移動(dòng)互聯(lián)網(wǎng)迎來(lái)了自己的又一個(gè)春天。4G網(wǎng)絡(luò)-作為上網(wǎng)速度快、延遲時(shí)間短、流量?jī)r(jià)格低的代名詞,不僅能實(shí)現(xiàn)移動(dòng)用戶(hù)的高速數(shù)據(jù)業(yè)務(wù),而且能較好地滿(mǎn)足移動(dòng)用戶(hù)高速和大容量的上網(wǎng)需求。再加上移動(dòng)終端的便攜性,將會(huì)有越來(lái)越多的用戶(hù)通過(guò)移動(dòng)終端接入互聯(lián)網(wǎng),那么對(duì)傳統(tǒng)互聯(lián)網(wǎng)用戶(hù)上網(wǎng)的行為,將會(huì)逐漸轉(zhuǎn)移到移動(dòng)互聯(lián)網(wǎng)上。這將為移動(dòng)互聯(lián)網(wǎng)的發(fā)展提供新的機(jī)遇,同時(shí)也會(huì)帶來(lái)新的挑戰(zhàn)。 21世紀(jì),不但是信息化的時(shí)代,也是講究時(shí)間效率的時(shí)代。在移動(dòng)運(yùn)營(yíng)商、電商風(fēng)靡的時(shí)代,誰(shuí)能搶在時(shí)間前頭,把有效精準(zhǔn)的服務(wù)和產(chǎn)品傳達(dá)給用戶(hù),為用戶(hù)提供個(gè)性化的業(yè)務(wù),從而保持其在同行業(yè)的領(lǐng)軍地位,以創(chuàng)造最大的收益。這是一個(gè)機(jī)遇,也是一個(gè)挑戰(zhàn)。那么對(duì)移動(dòng)用戶(hù)上網(wǎng)行為關(guān)鍵技術(shù)進(jìn)行研究以及對(duì)用戶(hù)上網(wǎng)行為進(jìn)行分析,了解用戶(hù)上網(wǎng)的特點(diǎn),從而為其提供高附加值精準(zhǔn)的個(gè)性化服務(wù)及高質(zhì)量的產(chǎn)品勢(shì)在必行。 本論文首先從分析移動(dòng)用戶(hù)上網(wǎng)行為的研究現(xiàn)狀、關(guān)鍵技術(shù)和研究方法著手,闡述了用戶(hù)行為分析的關(guān)鍵技術(shù),并通過(guò)分析已有的移動(dòng)用戶(hù)行為分析方法,提出了通過(guò)使用多維序列模式挖掘算法分析移動(dòng)用戶(hù)上網(wǎng)行為的方法。該方法是在挖掘出用戶(hù)頻繁瀏覽模式的前提下,考慮其它維的信息,如用戶(hù)IP、地域等,能更全面的反映用戶(hù)上網(wǎng)的特性和潛在的上網(wǎng)需求,能為網(wǎng)絡(luò)管理員進(jìn)行網(wǎng)站優(yōu)化與調(diào)整提供參考,能為運(yùn)營(yíng)商調(diào)整營(yíng)銷(xiāo)策略與模式提供依據(jù),從而又能為用戶(hù)提供強(qiáng)有效的服務(wù)及高品質(zhì)的產(chǎn)品提供保障。 接著,研究分析了目前已有的序列模式挖掘算法,如GSP、PrefixSpan、PSPM算法等和多維序列模式挖掘算法,如Seq-Dim、Seq-mdp等,并用實(shí)例分析比較了它們的優(yōu)缺點(diǎn)。在此研究分析的基礎(chǔ)上提出了本文改進(jìn)的3PC-MDSP算法,它與Seq-Dim、Seq-mdp等多維模式挖掘算法類(lèi)似,先挖掘出序列模式,然后在挖掘出序列模式的前提下,挖掘出多維模式。只是本算法在挖掘序列模式時(shí),采用本文改進(jìn)的3PC算法,即基于前綴位置投影位置元組比較的算法,在挖掘序列模式時(shí)不用構(gòu)造投影數(shù)據(jù)庫(kù),通過(guò)構(gòu)造前綴投影位置元組,根據(jù)元組投影位置為切入點(diǎn)掃描數(shù)據(jù)庫(kù),并通過(guò)比較不同元組的前綴位置與投影位置避免了頻繁項(xiàng)的重復(fù)掃描和挖掘,且比較元組數(shù)與最小支持度的大小,若小于最小支持度,則不用繼續(xù)挖掘,避免了對(duì)非頻繁模式挖掘時(shí)間的不必要的浪費(fèi),提高了算法的效率。并通過(guò)IBM數(shù)據(jù)生成器生成的數(shù)據(jù),驗(yàn)證了算法的有效性。 最后,設(shè)計(jì)并實(shí)現(xiàn)了移動(dòng)用戶(hù)上網(wǎng)行為分析系統(tǒng)。介紹了移動(dòng)用戶(hù)上網(wǎng)服務(wù)器端的Web日志格式,處理方法及數(shù)據(jù)存儲(chǔ)方法。并通過(guò)將改進(jìn)算法3PC-MDSP算法應(yīng)用到系統(tǒng)上對(duì)用戶(hù)上網(wǎng)行為進(jìn)行分析。
[Abstract]:With the Ministry of the issuance of 4G licenses, mobile Internet ushered in a spring.4G network as their Internet speed, short delay time, the flow of synonymous with low price, high speed data service can not only realize mobile users, and mobile users can better meet the high speed and large capacity needs of the Internet. Coupled with the mobile terminal, there will be more and more users through the mobile terminal access to the Internet, so the traditional Internet users online behavior, will be gradually transferred to the mobile Internet. This will provide new opportunities for the development of mobile Internet, but also brings new challenges.
In twenty-first Century, but also pay attention to the information age, time efficiency. In the era of mobile operators, electricity supplier in the era of popular, who can grab the time ahead, to convey accurate and effective products and services to users, to provide personalized services for users, so as to maintain its leading position in the same industry, to create the largest returns. This is an opportunity, but also a challenge. So the mobile Internet users research and Analysis on the behavior of Internet users are the key technology, understand the characteristics of Internet users, so as to provide high value-added personalized service accurate and high quality products for the imperative.
This paper first analyses the research status of mobile user behavior on the Internet, on key technologies and research methods, discusses the key technologies of user behavior analysis, and through the analysis of the existing mobile user behavior analysis method, proposed algorithm analysis method of mobile user behavior on the Internet through the use of sequential pattern mining. The method is to dig out the premise frequent user browsing patterns, consider the other dimension information, such as user IP, etc, can more fully reflect the characteristics of Internet users and potential needs of the Internet, network administrator can provide reference for network station optimization and adjustment, can provide the basis for the operator to adjust the marketing strategy and mode, which can provide service and high quality highly effective product for customers to provide protection.
Then, we analyze the present existing sequential pattern mining algorithms, such as GSP, PrefixSpan algorithm, PSPM algorithm and multidimensional pattern sequence mining such as Seq-Dim, Seq-mdp etc., and use examples to analyze and compare their advantages and disadvantages. Based on the analysis of the 3PC-MDSP algorithm proposed in this paper improved it, and Seq-Dim, Seq-mdp the multidimensional pattern mining algorithm is similar to mining sequence patterns, and then in the premise of mining sequential pattern mining, multidimensional model. But the algorithm in sequential pattern mining, this paper adopts the improved 3PC algorithm, namely comparison prefix position projection algorithm based on the position of the tuple, without constructing the projection database in sequence pattern mining by constructing the projection position, prefix tuples, tuple according to the projection position as a starting point to scan the database, and through the comparison of different prefix position and the projection position of the tuple to avoid the frequent Repeat scanning and mining items, and the number of tuples with minimum support size, if less than the minimum support, will not continue to dig, to avoid the non frequent pattern mining time of unnecessary waste, improve the efficiency of the algorithm. And through the IBM data generator generates the data, verify the validity of the algorithm.
Finally, a mobile user behavior analysis system is designed and implemented. The Web log format, processing method and data storage method of mobile user access server are introduced. By applying the improved algorithm 3PC-MDSP algorithm to the system, we analyze the user's online behavior.

【學(xué)位授予單位】：北京交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類(lèi)號(hào)】：TN929.5;TP311.13

【參考文獻(xiàn)】

相關(guān)期刊論文 前8條

1 肖志輝;;移動(dòng)互聯(lián)網(wǎng)研究綜述[J];電信科學(xué);2009年10期

2 劉威;程文青;;流媒體點(diǎn)播中用戶(hù)交互式行為建模[J];電子與信息學(xué)報(bào);2007年09期

3 張坤;朱揚(yáng)勇;;無(wú)重復(fù)投影數(shù)據(jù)庫(kù)掃描的序列模式挖掘算法[J];計(jì)算機(jī)研究與發(fā)展;2007年01期

4 繆裕青;吳孔玲;朱曉雁;蘇杰;;一種基于序列末項(xiàng)位置信息的序列模式挖掘算法[J];計(jì)算機(jī)應(yīng)用研究;2012年07期

5 張利軍;李戰(zhàn)懷;王淼;;基于位置信息的序列模式挖掘算法[J];計(jì)算機(jī)應(yīng)用研究;2009年02期

6 李滿(mǎn)天;尤佳莉;王勁林;王玲芳;;P2P網(wǎng)絡(luò)中基于K-means聚類(lèi)的用戶(hù)在線(xiàn)行為分析[J];微計(jì)算機(jī)應(yīng)用;2009年11期

7 葉進(jìn),張向利,張潤(rùn)蓮;基于數(shù)據(jù)挖掘的移動(dòng)客戶(hù)流失分析系統(tǒng)[J];計(jì)算機(jī)系統(tǒng)應(yīng)用;2005年02期

8 梁鵬,李兵,郭文君;移動(dòng)數(shù)據(jù)業(yè)務(wù)用戶(hù)行為模型研究[J];中興通訊技術(shù);2005年04期

，

本文編號(hào)：1747130