Search Results: 1 - 10 of 100 matches for " "
All listed articles are free for downloading (OA Articles)
Page 1 /100
Display every page Item
Granular association rules on two universes with four measures  [PDF]
Fan Min,Qinghua Hu,William Zhu
Computer Science , 2012,
Abstract: Relational association rules reveal patterns hide in multiple tables. Existing rules are usually evaluated through two measures, namely support and confidence. However, these two measures may not be enough to describe the strength of a rule. In this paper, we introduce granular association rules with four measures to reveal connections between granules in two universes, and propose three algorithms for rule mining. An example of such a rule might be "40% men like at least 30% kinds of alcohol; 45% customers are men and 6% products are alcohol." Here 45%, 6%, 40%, and 30% are the source coverage, the target coverage, the source confidence, and the target confidence, respectively. With these measures, our rules are semantically richer than existing ones. Three subtypes of rules are obtained through considering special requirements on the source/target confidence. Then we define a rule mining problem, and design a sandwich algorithm with different rule checking approaches for different subtypes. Experiments on a real world dataset show that the approaches dedicated to three subtypes are 2-3 orders of magnitudes faster than the one for the general case. A forward algorithm and a backward algorithm for one particular subtype can speed up the mining process further. This work opens a new research trend concerning relational association rule mining, granular computing and rough sets.
Cold-start recommendation through granular association rules  [PDF]
Fan Min,William Zhu
Computer Science , 2013,
Abstract: Recommender systems are popular in e-commerce as they suggest items of interest to users. Researchers have addressed the cold-start problem where either the user or the item is new. However, the situation with both new user and new item has seldom been considered. In this paper, we propose a cold-start recommendation approach to this situation based on granular association rules. Specifically, we provide a means for describing users and items through information granules, a means for generating association rules between users and items, and a means for recommending items to users using these rules. Experiments are undertaken on a publicly available dataset MovieLens. Results indicate that rule sets perform similarly on the training and the testing sets, and the appropriate setting of granule is essential to the application of granular association rules.
Mining top-k granular association rules for recommendation  [PDF]
Fan Min,William Zhu
Computer Science , 2013,
Abstract: Recommender systems are important for e-commerce companies as well as researchers. Recently, granular association rules have been proposed for cold-start recommendation. However, existing approaches reserve only globally strong rules; therefore some users may receive no recommendation at all. In this paper, we propose to mine the top-k granular association rules for each user. First we define three measures of granular association rules. These are the source coverage which measures the user granule size, the target coverage which measures the item granule size, and the confidence which measures the strength of the association. With the confidence measure, rules can be ranked according to their strength. Then we propose algorithms for training the recommender and suggesting items to each user. Experimental are undertaken on a publicly available data set MovieLens. Results indicate that the appropriate setting of granule can avoid over-fitting and at the same time, help obtaining high recommending accuracy.
Stochastic Mining of Quantitative Association Rules Using Multi Agent Systems  [PDF]
Zahra Karimi-Dehkordi,Mohmmadali Nematbakhsh,Ahmad. Baraani-Dastjerdi,Nasser Ghassem-Aghaee
ARPN Journal of Systems and Software , 2012,
Abstract: Discovering optimized intervals of numeric attributes in association rule mining has been recognized as an influential research problem over the last decade. There have been several stochastic optimization approaches such as evolutionary and swarm methods which try to find good intervals. One drawback of these approaches is sequential nature which requires multiple runs to find all rules. This paper presents multi agent architecture to find optimized rules simultaneously using adynamic priority schema. The Practical Swarm Optimization (PSO) Variant is modeled and implemented in JADEframework and tested with synthetic datasets. The results confirm finding the same sequential results in parallel.
Multi-Scaling Sampling: An Adaptive Sampling Method for Discovering Approximate Association Rules
Cai-Yan Jia,Xie-Ping Gao,

计算机科学技术学报 , 2005,
Abstract: One of the obstacles of the efficient association rule mining is the explosive expansion of data sets since it is costly or impossible to scan large databases, esp., for multiple times. A popular solution to improve the speed and scalability of the association rule mining is to do the algorithm on a random sample instead of the entire database. But how to effectively define and efficiently estimate the degree of error with respect to the outcome of the algorithm, and how to determine the sample size needed are entangling researches until now. In this paper, an effective and efficient algorithm is given based on the PAC (Probably Approximate Correct) learning theory to measure and estimate sample error. Then, a new adaptive, on-line, fast sampling strategy - multi-scaling sampling - is presented inspired by MRA (Multi-Resolution Analysis) and Shannon sampling theorem, for quickly obtaining acceptably approximate association rules at appropriate sample size. Both theoretical analysis and empirical study have showed that the sampling strategy can achieve a very good speed-accuracy trade-off.
Multi-objective Numeric Association Rules Mining via Ant Colony Optimization for Continuous Domains without Specifying Minimum Support and Minimum Confidence  [PDF]
Parisa Moslehi,Behrouz Minaei Bidgoli,Mahdi Nasiri,Afshin Salajegheh
International Journal of Computer Science Issues , 2011,
Abstract: Currently, all search algorithms which use discretization of numeric attributes for numeric association rule mining, work in the way that the original distribution of the numeric attributes will be lost. This issue leads to loss of information, so that the association rules which are generated through this process are not precise and accurate. Based on this fact, algorithms which can natively handle numeric attributes would be interesting. Since association rule mining can be considered as a multi-objective problem, rather than a single objective one, a new multi-objective algorithm for numeric association rule mining is presented in this paper, using Ant Colony Optimization for Continuous domains (ACOR). This algorithm mines numeric association rules without any need to specify minimum support and minimum confidence, in one step. In order to do this we modified ACOR for generating rules. The results show that we have more precise and accurate rules after applying this algorithm and the number of rules is more than the ones resulted from previous works.
Research and Application Multi-dimensional Association Rules Mining Based Artificial Immune System

ZHU Yu,ZHANG Hong,KONG Ling-dong,

计算机科学 , 2009,
Abstract: Association rules mining is very important in the application of data mining.At present,single-dimensional asso-ciation rules result have been matured,but the prominent combinatorial explosion problem of multi-dimensional association rules have not been solved perfectly so far.A method of mining multi-dimensional association rules was proposed based on artificial immune algorithm.This algorithm makes use of the immune memory characters,stores the asso-ciation rules in memory,and has faster speed of mining m...
Method based on rough set for mining multi-dimensional association rules

TAO Duo-xiu,LV Yue-jin,DENG Chun-yan,

计算机应用 , 2009,
Abstract: It is very time-consuming to discover association rules from the mass of data, and not all the rules are attractive to the user, so a lot of irrelevant information to the user's requirements may be generated when traditional mining methods are applied. In addition, most of the existing algorithms are for discovering one-dimensional association rules. Therefore, the authors defined a mining language which allowed users to specify items of interest to the association rules, as well as the parameters (for example, support, confidence, etc.). A method based on rough set theory for multi-dimensional association rules mining was also proposed, which dynamically generated frequent item sets and multi-dimensional association rules, and reduced the search space to generate frequent item sets. Finally, an example verifies the feasibility and effectiveness of the method.
Study on recommendation algorithm based on multi-level association rules

YU Xiao-peng,

计算机应用 , 2007,
Abstract: A model-based recommendation algorithm was proposed, which uses Multi-level Association Rules (MAR) to alleviate those problems about data sparseness and scalability of the recent recommendation algorithm. In this algorithm, a model for preference prediction was built by using multi-level association rule mining, which is used to compute preferences for items. The experimental results show that performance of the algorithm is superior to other methods.
Mining of Multi-Relational Association Rules

HE Jun,LIU Hong-Yan,DU Xiao-Yong,

软件学报 , 2007,
Abstract: Association rule mining is one of the most important and basic technique in data mining,which has been studied extensively and has a wide range of applications.However,as traditional data mining algorithms usually only focus on analyzing data organized in single table,applying these algorithms in multi-relational data environment will result in many problems.This paper summarizes these problems,proposes a framework for the mining of multi-relational association rule,and gives a definition of the mining task.After classifying the existing work into two categories,it describes the main techniques used in several typical algorithms,and it also makes comparison and analysis among them.Finally,it points out some issues unsolved and some future further research work in this area.
Page 1 /100
Display every page Item

Copyright © 2008-2017 Open Access Library. All rights reserved.