Search

For that the existing community discovery algorithms of multi-layered social network (MSN) are lack and their results are crude, this paper presented cross-layer edge differential clustering coefficient(CLEDCC) algorithm in MSN based on edge clustering. This algorithm considered the strength of the relationships about any two nodes and their neighbors in each layer network, and respectively simulated for artificial sparse network, dense network and real datasets. Experimental results show that CLEDCC algorithm can effectively avoid the problem of parameter uncertainty, which is more accurate than the cross-layer edge clustering coefficient CLECC) algorithm in community discovery. [ABSTRACT FROM AUTHOR]

Finite-domain representations (FDR) is one of the best methods to compress the size of state space, but it requires building a logic program for each initial state in grounding stage. Unfortunately the number of the possible initial state in conformant planning is always huge. So there will be heavy time and space overhead and even leads to memory overflow. In order to improve the efficiency of the algorithm, this paper designed a new method named IFDR using a single logic program to translate the PPDDL task into grounded one. A logic program consists of a set of facts and a set of rules. The set of facts of the logic program was formed by the atoms in all the possible initial state, and the set of rules was derived from the axiom and operator definitions. The outcomes of comparative experiments validate the accuracy and effectiveness of the new algorithm. [ABSTRACT FROM AUTHOR]

After analyzing the sparsity of traditional collaborative filtering algorithm and the merit and demerit of current solutions, this paper proposed a collaborative filtering recommendation algorithm based on user characteristics and item attributes which was mainly based on the item-based collaborative algorithm. It predicted the unrated items by analyzing different users' interests to various attributes of items and integrating the attributes of rated items to reduce the sparsity of data sets, and then to improve the accuracy of items' similarity calculation. The experimental results based on MovieLens data set show that the sparsity of extreme data sets can be reduced effectively, to a certain extent the proposed algorithm alleviates the cold starting problem in collaborative filtering algorithm and achieves better prediction accuracy. [ABSTRACT FROM AUTHOR]