Skip to main content

Table 1 Comparison of state of the art methods

From: Feature-enriched author ranking in incomplete networks

Author-level Method Initialisation: R(Ni) Score term: ST(Ni)
  RLPR (Ding 2009) \(\frac {1}{|\mathcal {A}|}\) \((1-q) \sum \limits _{(A_{i'} \rightarrow A_{i}, P_{j})}\frac {S(A_{i'}) \times w({A_{i'}\rightarrow A_{i}}, P_{j})}{w_{out}(A_{i'})}\)
  SARA (Radicchi et al. 2009) \(\frac {\sum \limits _{(P_{j} \in \mathcal {P}_{A_{i}})}\frac {1}{|\mathcal {A}_{P_{j}}|}}{\sum \limits _{(A_{i'} \in \mathcal {A})}\sum \limits _{(P_{j} \in \mathcal {P}_{A_{i'}})}\frac {1}{|\mathcal {A}_{P_{j}}|}}\)  
  ALEF (West et al. 2013) \(\frac {|\mathcal {P}_{A_{i}}|}{|\mathcal {P}|}\)  
  SCEAS (Sidiropoulos and Manolopoulos 2006) \(\frac {1}{|\mathcal {A}|}\) \( \frac {(1-q)}{a} \sum \limits _{(A_{i'} \rightarrow A_{i}, P_{j})}\frac {(S(A_{i'})+b) \times w({A_{i'}\rightarrow A_{i}}, P_{j})}{w_{out}(A_{i'})}\)
Paper-level YetRank (Hwang et al. 2010) \(v(P_{i}) \times \frac {e^{\frac {-\delta (P_{i})}{\tau }}}{\tau }\) \((1-q) \sum \limits _{(P_{i'}\rightarrow P_{i})} \frac {S(P_{i'})\times R(P_{i})}{r_{out}(P_{i'})} \)
  NewRank (Dunaiski and Visser 2012) \(e^{\frac {-\delta (P_{i})}{\tau }}\)  
  1. Ni represents a node in the network, i.e., Ni=Ai in author-level networks, and Ni=Pi in paper-level networks. Score diffusion S(Ni) is equal to ST(Ni)+RR(Ni)+DN(Ni). For all methods, RR(Ni)=q×R(Ni) and DN(Ni)=(1−qR(Ni), thus we omit them from the table