Identifying Network Public Opinion Leaders Based on Markov Logic Networks

Box 3

//Waterarmy

repeat(a1,+a2)=>act(a1,+a3)

//Leader

totalreplynum(a1,+a2)^∧totalpostnum(a1,+a3)^∧totalbereplynum(a1,+a4)=>act(a1,+a5)

post(a1,a2)^∧replynumofpost(a2,+a3)^∧clicknumofpost(a2,+a4)=>act(a1,+a5)

//Try to take advantage of ralational data

act(a1,+a2)^∧follow(a1,a3)=>act(a3,+a4)

act(a1,+a2)^∧reply(a1,a3)=>act(a3,+a4)

//Passer

//totalreplynum(a1,+a2)^∧totalpostnum(a1,+a3)^∧totalbereplynum(a1,+a4)=>act(a1,+a5)