【摘要】：隨著微博這一新媒體的發(fā)展,微博營銷迅速興起,成為各行各業(yè)關(guān)注的焦點(diǎn)。在微博這一新興社交媒體平臺(tái)上,營銷行為不僅引起了廣大媒體人的關(guān)注,同時(shí)也得到了學(xué)術(shù)界的青睞,微博營銷行為所產(chǎn)生的海量數(shù)據(jù),正是發(fā)揮數(shù)據(jù)挖掘技術(shù)優(yōu)勢(shì)的地方。因此,本文通過研究數(shù)據(jù)挖掘方法,借助相關(guān)算法對(duì)微博營銷行為特征進(jìn)行挖掘分析。 本文主要研究了新浪微博中的電影微博營銷行為及相關(guān)用戶偏好特征。文章首先提出了一種基于新浪微博文本內(nèi)容的話題提取算法——基于模糊C均值(FCM)聚類算法的話題提取算法(TEBF)。該算法在FCM算法中隸屬度(membership)這一概念的基礎(chǔ)上提出了M-membership這一指標(biāo),實(shí)驗(yàn)證明該指標(biāo)可以很好地衡量模糊集合中詞語的重要性,最終實(shí)現(xiàn)以模糊集合的形式描述話題,并通過對(duì)比實(shí)驗(yàn)說明了TEBF算法話題提取結(jié)果的有效性。然后,文章結(jié)合微博營銷行為和用戶偏好分析的具體需求,基于話題提取算法和分類算法,提出了營銷行為分析模型和用戶偏好分析模型,從微博內(nèi)容和用戶信息兩個(gè)方面對(duì)微博營銷行為及用戶偏好進(jìn)行了較為全面系統(tǒng)的分析。 為了對(duì)分析結(jié)果進(jìn)行簡單直觀的展示,本文設(shè)計(jì)開發(fā)了“WMA”微博營銷行為分析系統(tǒng),以期為更多的用戶提供一個(gè)方便有效的微博營銷行為及用戶偏好分析平臺(tái),同時(shí)也證明本文的研究具有一定的實(shí)用價(jià)值。
[Abstract]:With the development of Weibo, Weibo marketing has become the focus of various industries. On the emerging social media platform, Weibo, the marketing behavior not only attracts the attention of the media, but also gets the favor of the academic circles. The massive data produced by the Weibo marketing behavior is the place where the advantages of data mining technology can be brought into play. Therefore, this paper through the study of data mining methods, with the help of related algorithms to analyze the characteristics of Weibo marketing behavior. This paper mainly studies the marketing behavior of movie Weibo and the characteristics of user preference in Sina Weibo. In this paper, we first propose a topic extraction algorithm based on Sina Weibo text content-topic extraction algorithm (TEBF). Based on fuzzy C-means clustering algorithm. Based on the concept of membership degree (membership) in FCM algorithm, this algorithm puts forward M-membership index. The experiment proves that this index can well measure the importance of words in fuzzy set, and finally describe the topic in the form of fuzzy set. The validity of the topic extraction results of the TEBF algorithm is proved by a comparative experiment. Then, combined with the specific requirements of Weibo marketing behavior and user preference analysis, based on topic extraction algorithm and classification algorithm, this paper proposes a marketing behavior analysis model and a user preference analysis model. This paper makes a comprehensive and systematic analysis of Weibo marketing behavior and user preference from the aspects of Weibo content and user information. In order to display the analysis results simply and intuitively, this paper designs and develops the "WMA" Weibo marketing behavior analysis system, in order to provide more users with a convenient and effective Weibo marketing behavior and user preference analysis platform. At the same time, it is proved that the research in this paper has certain practical value.
【學(xué)位授予單位】：北京郵電大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TP391.1

【參考文獻(xiàn)】

相關(guān)期刊論文 前3條

1 葛進(jìn)平;鄒立清;;電影微博立體營銷策略探討[J];當(dāng)代電影;2012年02期

2 徐戈;王厚峰;;自然語言處理中主題模型的發(fā)展[J];計(jì)算機(jī)學(xué)報(bào);2011年08期

3 張保富;施化吉;馬素琴;;基于TFIDF文本特征加權(quán)方法的改進(jìn)研究[J];計(jì)算機(jī)應(yīng)用與軟件;2011年02期

，

本文編號(hào)：2453227