Pi-Lian He

The research of improved association rules mining Apriori algorithm

This paper points out the bottleneck of classical Aprioris algorithm, presents an improved association rule mining algorithm. The new algorithm is based on reducing the times of scanning candidate sets and using hash tree to store candidate itemsets. According to the running result of the algorithm, the processing time of mining is decreased and the efficiency of algorithm has improved.1

A high efficient AprioriTid algorithm for mining association rule

Mining association rule is one of the common forms in data mining, in which the critical problem is to gain the frequent itemsets efficiently. The classical Apriori and AprioriTid algorithm, which are used to construct the frequent itemset, are analyzed in this paper. Author finds out that there too many data due to those items repeatedly saved in the AprioriTid algorithm. On the basis of analysis, we give a theorem of the itemset whose support is less than minsup in C/sub k-1/ is useless in C/sub k-1/. Then, HEA algorithm based on the theorem is offered. The experiments show that the new algorithm is more effective in decreasing data size and execution times than AprioriTid algorithm.

An improved term weighting scheme for vector space model

Document representation has been the fundamental issue in the information retrieval (IR). However, the traditional vector space model (VSM) has data sparseness phenomena on the representation of document vectors, and cannot well discriminate the expression competence to the document content of indexing terms in different positions. This paper proposes an improved term weighting method by introducing information gain of terms while taking above factors into account. The theoretical analysis and experimental results show that the new scheme improves the performance of VSM in IR in terms of higher recall and precision.

Research of semantic caching for LDQ in mobile network

Location-dependent query is becoming very popular in mobile environments. To improve system performance, many semantic cache models are proposed. In this paper, we first define a new structure of semantic segments index. Then we analyze the detailed query processing and FAR algorithm. For overcoming the shortage of predicting future location in FAR, we propose a new improved algorithm (RBF-FAR) as replacement policy, which use RBFNN to predicate next location instead of velocity in FAR. Using this new replacement policy, we propose a new model for LDQ semantic cache. The experiment results show that new model, on the basis of RBF-FAR, is more flexible and effective at reducing average response time network-load used in LDQ than FAR model

Algorithm of documents clustering based on minimum spanning tree

This paper puts forward a method of document clustering based on minimum spanning tree (MST) in vector space model (VSM). This algorithm adopts classical VSM and combines with the method of MST in graph theory. The quality and performance of document clustering are higher than other traditional clustering methods.

A Local Computing-Based Hierarchical Clustering Algorithm Building Density Trees

A new kind of clustering algorithm called LOCHDET (LOcal Computing-based Hierarchical clustering algorithm building DEnsity Trees) is presented in this paper. LOCHDET generates a density tree for each potential cluster according to its local density distribution. Each cluster is regarded as a tight coupling structure. Those “closer” clusters are merged if some conditio are satisfied. In order to reduce the cost time, a local computing technology is introduced. LOCHDET has a wide range of parameter settings, preferable accuracy in discovering clusters with arbitrary shape, good ability of processing noise data sets and weak sensitivity to input parameters by generalizing density-based, hierarchical, and locality-based methods. The results of our experiments confirm these mentioned above.

Dynamic clustering analysis of documents based on cluster centroids

This paper puts forward a method of documents dynamic clustering based on cluster centroids. First, the documents are modeled as elements in vector space and clustered based on some cluster centroids. Then the clustering results are dynamically adjusted based on the principle of group modification. Finally, the experiment results of documents clustering are contrasted, which show dynamic clustering can improve the performance of documents clustering to a certain extent.

Predictive control based on neural network for nonlinear system with time-delay

Proposes a predictive control strategy based on an improved a BP neural network in order to compensate real time control in a nonlinear system with time-delay. The scheme is applied and the simulation result for process control is presented to illustrate the possible benefits with the given concept.

Applying RBF Network to Predict Location in Mobile Network

In mobile network, quality of service (Qos) is difficultly guaranteed for the particularity of mobile network. If the system knows, prior to the mobile subscriber movement, the exact trajectory it will follow, the Qos can be guaranteed. Thus, location prediction is the key issue to provide quality of service to mobile subscriber. In the present paper, RBF Network of Neural Network techniques were used to predict the mobile users next location based on his current location as well as time. The software matlab 6.5 was used to confirm the parameters of RBF network, and to same training data, makes the detailed contrast with resilient propagation BP and BP in learning time and steps of learning. Experiment results show that predicted locations with RBF are more effective and accurate than resilient BP.

Improving Searching Performance Based on Semantic Correlativity in Peer to Peer Network

Most existing peer-to-peer (P2P) systems support only title-based searches, which can not satisfy the content searches. In this paper, we proposed a semantic correlativity model which can support semantic content-based searches. Firstly, using VSM to represent content and using KNN algorithm to implement self- clustering. Secondly, based on framework, accessing to compute semantic similarity, SCRA policy is proposed to improve routing performance with prefetch technology. By this model, routing overhead can be greatly reduced. At last, preliminary simulation results show that SCRA achieves a great routing performance over the previous algorithms.