Negative Generator Border for Effective Pattern Maintenance

In this paper, we study the maintenance of frequent patterns in the context of the generator representation. The generator representation is a concise and lossless representation of frequent patterns. We effectively maintain the generator representation by systematically expanding its Negative Generator Border. According to our literature review, no prior work has studied the maintenance of the generator representation. To illustrate the proposed maintenance idea, a new algorithm is developed to maintain the generator representation for support threshold adjustment. Our experimental results show that the proposed algorithm is significantly faster than other state-of-the-art algorithms. This proposed maintenance idea can also be extended to other representations of frequent patterns as demonstrated in this paper.

• Type: Conference
• Year: 2008
• Where: ADMA
• Authors: Mengling Feng, Jinyan Li, Limsoon Wong, Yap-Peng Tan
• Keywords: ADMA 2008, Data Mining, Frequent Patterns, Generator Representation, Negative Generator Border