Compact Weighted Class Association Rule Mining Using Information Gain

Ibrahim, S. P. Syed; Chandran, Karthik

doi:10.5121/ijdkp.2011.1601

Cited by 14 publications

(7 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Weight of each item is derived by using HITS model and WARM is performed [14]. This model produces quality oriented rules of lesser number to improve the accuracy of classification.…”

Section: Related Workmentioning

confidence: 99%

Recent and Frequent Informative Pages from Web Logs by Weighted Association Rule Mining

Malarvizhi.¹

2019

IJMECS

View full text Add to dashboard Cite

Web Usage Mining provides efficient ways of mining the web logs for knowing the user's behavioral patterns. Existing literature have discussed about mining frequent pages of web logs by different means. Instead of mining all the frequently visited pages, if the criterion for mining frequent pages is based on a weighted setting then the compilation time and storage space would reduce. Hence in the proposed work, mining is performed by assigning weights to web pages based on two criteria. One is the time dwelled by a visitor on a particular page and the other is based on recent access of those pages. The proposed Weighted Window Tree (WWT) method performs Weighted Association Rule mining (WARM) for discovering the recently accessed frequent pages from web logs where the user has dwelled for more time and hence proves that these pages are more informative. WARM's significance is in page weight assignment for targeting essential pages which has an advantage of mining lesser quality rules.

show abstract

“…Weight of each item is derived by using HITS model and WARM is performed [14]. This model produces quality oriented rules of lesser number to improve the accuracy of classification.…”

Section: Related Workmentioning

confidence: 99%

Recent and Frequent Informative Pages from Web Logs by Weighted Association Rule Mining

Malarvizhi.¹

2019

IJMECS

View full text Add to dashboard Cite

show abstract

“…Identifying and extracting patterns and regularities in massive data repositories has been a focused theme in data mining research for almost over a decade. Substantial progress continues to be made in this context, specially in the tasks of frequent itemset mining [15] and association rule mining [7], [8], [10], [16], [17]. Temporal pattern discovery is a very promising extension to this ongoing research theme because it substantially broadens the scope of data analysis by supporting the discovery of patterns and regularities that are time-dependent.…”

Section: Introductionmentioning

confidence: 99%

Detection of Calendar-Based Periodicities of Interval-Based Temporal Patterns

Dutta¹

2012

IJDKP

View full text Add to dashboard Cite

show abstract

“…In [7] and [8] the authors proposed information gain attribute based approach for associative classification where high informative attribute is chosen for rule generation. Based on this idea the authors of [9] and [10] proposed information gain based weighted associative classification methods. In [11] the authors proposed associative classification method using genetic network programming.…”

Section: Introductionmentioning

confidence: 99%

LACI: Lazy Associative Classification using Information Gain

Ibrahim¹,

Chandran²,

Kanthasamy³

2012

IJET

View full text Add to dashboard Cite

Associative classification method applies association rule mining technique in classification and achieves higher classification accuracy. However, it is a known fact that associative classification typically yields a large number of rules, from which a set of high quality rules are chosen to construct an efficient classifier. Hence, generating, ranking and selecting a small subset of high-quality rules without jeopardizing the classification accuracy is of prime importance but indeed a challenging task. Lazy learning associative classification method eliminates the need of constructing the classifier but suffers with high computation cost. This paper proposes lazy associative classification using Information gain where, the system first chooses the Information gained attribute from the training sample and computes highest subset probability and then it directly predicts the class label. This proposed method not only reduces the computation cost but also improves the classification accuracy. Experimental result shows that the proposed system outperforms the traditional associative classification methods and the existing lazy associative classification method.

show abstract

Compact Weighted Class Association Rule Mining Using Information Gain

Cited by 14 publications

References 22 publications

Recent and Frequent Informative Pages from Web Logs by Weighted Association Rule Mining

Recent and Frequent Informative Pages from Web Logs by Weighted Association Rule Mining

Detection of Calendar-Based Periodicities of Interval-Based Temporal Patterns

LACI: Lazy Associative Classification using Information Gain

Contact Info

Product

Resources

About