Optimization of quality measures in association rule mining: an empirical study

Luna, José María; Ondra, M.; Fardoun, Habib M.; Ventura, Sebastián

doi:10.2991/ijcis.2018.25905182

Cited by 24 publications

(14 citation statements)

References 23 publications

(38 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The frequency of occurrence of P in Ω, also known as support(P ) [26], is defined as the number of bags in which P satisfies at least one transaction, i.e. support(P ) =…”

Section: A Single Level Of Ambiguitymentioning

confidence: 99%

Extracting User-Centric Knowledge on Two Different Spaces: Concepts and Records

2020

View full text Add to dashboard Cite

The growing demand for eliciting useful knowledge from data calls for techniques that can discover insights (in the form of patterns) that users need. Methodologies for describing intrinsic and relevant properties of data through the extraction of useful patterns, however, work on fixed input data, and the data representation, therefore, constrains the discovered insights. In this regard, this paper aims at providing foundations to make the descriptive knowledge that is extracted by pattern mining more user-centric by relying on flexible data structures defined on two different perspectives: concepts and data records. In this sense, items in data can be grouped into abstract terms through subjective hierarchies of concepts, whereas data records can also be organized based on the users' subjective perspective. A series of easy-to-follow toy examples are considered for each of the two perspectives to demonstrate the usefulness and necessity of the proposed foundations in pattern mining. Finally, aiming at experimentally testing whether classical pattern mining algorithms can be adapted to such flexible data structures, the experimental analysis comprises different methodologies, including exhaustive search, random search, and evolutionary approaches. All these approaches are based on well-known and widely recognized techniques to demonstrate the usefulness of the provided foundations for future research works and more efficient and specifically designed algorithms. Obtained insights demonstrate the importance of working with subjectivity: an item is a type of soda but belongs to a pack, including two or more soda types.

show abstract

“…The frequency of occurrence of P in Ω, also known as support(P ) [26], is defined as the number of bags in which P satisfies at least one transaction, i.e. support(P ) =…”

Section: A Single Level Of Ambiguitymentioning

confidence: 99%

Extracting User-Centric Knowledge on Two Different Spaces: Concepts and Records

2020

View full text Add to dashboard Cite

show abstract

“…who wants to get actionable insights. In this regard, interestingness quality measures (Luna, Ondra, Fardoun, & Ventura, 2018) can be used to filter and/or rank the output. These measures are divided into objective or data-driven (statistical and structural properties of data) and subjective or user-driven (user's preferences and goals).…”

Section: Definition 2 (Frequency Of a Pattern)mentioning

confidence: 99%

Frequent itemset mining: A 25 years review

Luna

Fournier-Viger

Ventura

2019

WIREs Data Min & Knowl

170

View full text Add to dashboard Cite

Frequent itemset mining (FIM) is an essential task within data analysis since it is responsible for extracting frequently occurring events, patterns, or items in data. Insights from such pattern analysis offer important benefits in decision‐making processes. However, algorithmic solutions for mining such kind of patterns are not straightforward since the computational complexity exponentially increases with the number of items in data. This issue, together with the significant memory consumption that is present in the mining process, makes it necessary to propose extremely efficient solutions. Since the FIM problem was first described in the early 1990s, multiple solutions have been proposed by considering centralized systems as well as parallel (shared or nonshared memory) architectures. Solutions can also be divided into exhaustive search and nonexhaustive search models. Many of such approaches are extensions of other solutions and it is therefore necessary to analyze how this task has been considered during the last decades. This article is categorized under: Algorithmic Development > Association Rules Technologies > Association Rules

show abstract

“…The effectiveness of the association rule algorithm requires quantitative indicators for evaluation. As is shown in Equations (10) and (11), accuracy and lift [34]- [36] are used to evaluate R-FP-growth algorithm proposed in this paper. Here, lift appear as one of the two most promising quality measures to be used as metrics in association rule mining problem [36], which has been widely used to evaluate association rule mining results [37]- [40].…”

Section: B Evaluation Metricsmentioning

confidence: 99%

“…As is shown in Equations (10) and (11), accuracy and lift [34]- [36] are used to evaluate R-FP-growth algorithm proposed in this paper. Here, lift appear as one of the two most promising quality measures to be used as metrics in association rule mining problem [36], which has been widely used to evaluate association rule mining results [37]- [40]. accuracy = N true N all (10) lift(I X → I Y ) = P(I X I Y ) P(I X )P(I Y ) (11) In Equation 10: Accuracy indicates the accuracy of the association rule algorithm, i.e., the ratio of the number of valid rules N true to the total number of rules N all in the association rule mining result, which uses lift to determine whether a certain rule is valid.…”

Section: B Evaluation Metricsmentioning

confidence: 99%

Association Rule Mining Method Based on the Similarity Metric of Tuple-Relation in Indoor Environment

et al. 2020

View full text Add to dashboard Cite

Association rules can detect the association pattern between POIs (point of interest) and serve the application of indoor location. In this paper, a new index, tuple-relation, is defined, which reflects the association strength between POI sets in indoor environment. This index considers the potential association information such as spatial and semantic information between indoor POI sets. On this basis, a new R-FP-growth (tuple-relation frequent pattern growth) algorithm for mining association rules in indoor environment is proposed, which makes comprehensive use of the co-occurrence probability, conditional probability, and multiple potential association information among POI sets, to form a new support-confidence-relation constraint framework and to improve the quality and application value of mining results. Experiments are performed, using real Wi-Fi positioning trajectory data from a shopping mall. Experimental results show that the tuple-relation calculation method based on cosine similarity has the best effect, with an accuracy of 87%, and 19% higher than that of the traditional FP-growth algorithm. INDEX TERMS Association rule, data mining, indoor trajectory, network embedding.

show abstract

Optimization of quality measures in association rule mining: an empirical study

Cited by 24 publications

References 23 publications

Extracting User-Centric Knowledge on Two Different Spaces: Concepts and Records

Extracting User-Centric Knowledge on Two Different Spaces: Concepts and Records

Frequent itemset mining: A 25 years review

Association Rule Mining Method Based on the Similarity Metric of Tuple-Relation in Indoor Environment

Contact Info

Product

Resources

About