计算机工程与设计
計算機工程與設計
계산궤공정여설계
COMPUTER ENGINEERING AND DESIGN
2015年
6期
1620-1625
,共6页
网络舆情%博客话题%热点发现%链接%萤火虫算法%聚类%簇结构
網絡輿情%博客話題%熱點髮現%鏈接%螢火蟲算法%聚類%簇結構
망락여정%박객화제%열점발현%련접%형화충산법%취류%족결구
Internet public opinion%Blog topic%hotspot discovery%link%firefly algorithm%clustering%cluster structure
针对已有的基于链接分析的热点发现方法存在准确度较低、易受作弊链接影响、易产生主题漂移现象等问题，利用复杂网络簇结构具有高度主题相关的特点，提出一种融合应用链接分析和萤火虫算法聚类博文的热点话题发现算法。以博文页面为节点，与博文内容相同或相关的链接作为边，根据博文及博主的相关属性，综合评定页面权重，建立博客话题模型；运用萤火虫算法对博文进行聚类获得聚类中心，按页面权重将聚类中心从大到小排序，形成热点话题热度排行。实验结果表明，该方法能够发现精度更高、数量更多的博客热点话题。
針對已有的基于鏈接分析的熱點髮現方法存在準確度較低、易受作弊鏈接影響、易產生主題漂移現象等問題，利用複雜網絡簇結構具有高度主題相關的特點，提齣一種融閤應用鏈接分析和螢火蟲算法聚類博文的熱點話題髮現算法。以博文頁麵為節點，與博文內容相同或相關的鏈接作為邊，根據博文及博主的相關屬性，綜閤評定頁麵權重，建立博客話題模型；運用螢火蟲算法對博文進行聚類穫得聚類中心，按頁麵權重將聚類中心從大到小排序，形成熱點話題熱度排行。實驗結果錶明，該方法能夠髮現精度更高、數量更多的博客熱點話題。
침대이유적기우련접분석적열점발현방법존재준학도교저、역수작폐련접영향、역산생주제표이현상등문제，이용복잡망락족결구구유고도주제상관적특점，제출일충융합응용련접분석화형화충산법취류박문적열점화제발현산법。이박문혈면위절점，여박문내용상동혹상관적련접작위변，근거박문급박주적상관속성，종합평정혈면권중，건립박객화제모형；운용형화충산법대박문진행취류획득취류중심，안혈면권중장취류중심종대도소배서，형성열점화제열도배행。실험결과표명，해방법능구발현정도경고、수량경다적박객열점화제。
To address the problem that the existing hotspot topics finding methods use link analysis to obtain the results ,which results in lower precision ,vulnerability to cheating links and emerging topic drift phenomenon ,a hotspot topic discovery algo‐rithm using link analysis and firefly algorithm to cluster Blog articles was proposed based on the feature of highly relevant subject of cluster structures in complex networks .Blog article pages and the links of the same or relative Blog articles were represented by nodes and edges respectively ,and the weights of Blog article pages were evaluated using the relative attributes of Blog articles and Bloggers ,and then the Blog topic model was established .Firefly algorithm was used to cluster the Blog articles to form the cluster centers ,and these cluster centers were sorted using the weights of the Blog article pages to obtain the ranking of hotspot topics .The experimental results show that the proposed method can find more Blog hotspot topics with higher precision .
