【摘要】：社區(qū)發(fā)現(xiàn)是社會(huì)網(wǎng)絡(luò)研究的熱點(diǎn)問題,綜合利用社會(huì)網(wǎng)絡(luò)中不同對象間的異質(zhì)信息,可以更加有效地挖掘網(wǎng)絡(luò)中的社區(qū)結(jié)構(gòu).針對傳統(tǒng)的社區(qū)發(fā)現(xiàn)方法無法有效地利用異質(zhì)信息的問題,本文提出了一種基于語義路徑的異質(zhì)網(wǎng)絡(luò)社區(qū)發(fā)現(xiàn)方法,該方法首先定義網(wǎng)絡(luò)中的語義路徑,通過語義路徑來衡量不同類型對象間的異質(zhì)信息相似度,然后以此構(gòu)造可靠性矩陣,作為半監(jiān)督非負(fù)矩陣分解的正則化約束項(xiàng),進(jìn)而實(shí)現(xiàn)異質(zhì)網(wǎng)絡(luò)的社區(qū)劃分.在真實(shí)數(shù)據(jù)集上的實(shí)驗(yàn)結(jié)果表明,所提出的方法能夠更準(zhǔn)確地發(fā)現(xiàn)異質(zhì)網(wǎng)絡(luò)中的社區(qū)結(jié)構(gòu).
[Abstract]:Community discovery is a hot topic in social network research. Using heterogeneous information among different objects in social network can effectively excavate the community structure in social network. In order to solve the problem that traditional community discovery methods can not effectively utilize heterogeneous information, this paper proposes a semantic path-based heterogeneous network community discovery method, which first defines semantic paths in the network. The similarity of heterogeneous information between different types of objects is measured by semantic paths, and then the reliability matrix is constructed as a regularization constraint for semi-supervised non-negative matrix decomposition, and then the community partition of heterogeneous networks is realized. Experimental results on real data sets show that the proposed method can more accurately detect the community structure in heterogeneous networks.
【作者單位】： 國家數(shù)字交換系統(tǒng)工程技術(shù)研究中心;
【基金】：國家科技支撐計(jì)劃(No.2014BAH30B01)
【分類號】：TP393.09
，

本文編號：2355951