基于種子節(jié)點(diǎn)選擇的重疊社區(qū)發(fā)現(xiàn)算法
發(fā)布時(shí)間:2019-01-03 08:07
【摘要】:針對(duì)目前從局部社區(qū)擴(kuò)展成全局社區(qū)時(shí)有關(guān)算法的種子節(jié)點(diǎn)選擇不合理的情形,提出了一種基于種子節(jié)點(diǎn)選擇的重疊社區(qū)發(fā)現(xiàn)算法。首先根據(jù)影響力函數(shù)找出局部影響力最大的節(jié)點(diǎn),由這些節(jié)點(diǎn)構(gòu)成的種子集合較好地分布在整個(gè)網(wǎng)絡(luò)中,然后以這些種子點(diǎn)構(gòu)造初始社區(qū),根據(jù)設(shè)定的吸引度函數(shù)選擇性地添加節(jié)點(diǎn)來進(jìn)行社區(qū)擴(kuò)展。實(shí)驗(yàn)結(jié)果表明,該算法在真實(shí)網(wǎng)絡(luò)上進(jìn)行測(cè)試時(shí)能夠有效地挖掘網(wǎng)絡(luò)中的重疊社區(qū)。
[Abstract]:An overlapping community discovery algorithm based on seed node selection is proposed in view of the unreasonable selection of seed nodes when the algorithm is extended from a local community to a global community. Firstly, the nodes with the most local influence are found according to the influence function, and the seed set composed of these nodes is distributed well in the whole network, and then the initial community is constructed from these seed points. Selectively add nodes according to the set attractiveness function to extend the community. Experimental results show that the algorithm can effectively mine overlapping communities in real networks.
【作者單位】: 中國(guó)人民大學(xué)信息學(xué)院;淮陰師范學(xué)院計(jì)算機(jī)科學(xué)與技術(shù)學(xué)院;
【基金】:國(guó)家自然科學(xué)基金資助項(xiàng)目(71271211,71531012) 北京市自然科學(xué)基金資助項(xiàng)目(4132067) 中國(guó)人民大學(xué)品牌計(jì)劃資助項(xiàng)目(10XNI029)
【分類號(hào)】:TP301.6
本文編號(hào):2399100
[Abstract]:An overlapping community discovery algorithm based on seed node selection is proposed in view of the unreasonable selection of seed nodes when the algorithm is extended from a local community to a global community. Firstly, the nodes with the most local influence are found according to the influence function, and the seed set composed of these nodes is distributed well in the whole network, and then the initial community is constructed from these seed points. Selectively add nodes according to the set attractiveness function to extend the community. Experimental results show that the algorithm can effectively mine overlapping communities in real networks.
【作者單位】: 中國(guó)人民大學(xué)信息學(xué)院;淮陰師范學(xué)院計(jì)算機(jī)科學(xué)與技術(shù)學(xué)院;
【基金】:國(guó)家自然科學(xué)基金資助項(xiàng)目(71271211,71531012) 北京市自然科學(xué)基金資助項(xiàng)目(4132067) 中國(guó)人民大學(xué)品牌計(jì)劃資助項(xiàng)目(10XNI029)
【分類號(hào)】:TP301.6
【相似文獻(xiàn)】
相關(guān)期刊論文 前3條
1 鄭重;王意潔;馬行空;;SubCounter:一種基于語義簇聚的節(jié)點(diǎn)子集規(guī)模估計(jì)方法[J];計(jì)算機(jī)工程與科學(xué);2012年02期
2 王甜;葉德建;;一種基于全局AS視圖的BitTorrent節(jié)點(diǎn)選擇算法[J];計(jì)算機(jī)應(yīng)用與軟件;2010年11期
3 宋偉;梁霖;李鋒;;新型影響力最大化算法[J];電視技術(shù);2014年15期
相關(guān)碩士學(xué)位論文 前2條
1 陳琳(Linda Chen);基于行為傳播的初始節(jié)點(diǎn)選擇[D];南京郵電大學(xué);2015年
2 朱丹丹;考慮節(jié)點(diǎn)屬性汽車導(dǎo)航系統(tǒng)路徑規(guī)劃問題的研究[D];天津理工大學(xué);2014年
,本文編號(hào):2399100
本文鏈接:http://sikaile.net/kejilunwen/ruanjiangongchenglunwen/2399100.html
最近更新
教材專著