Discovery & effective use of quality association rules in multi-level datasets

In today’s electronic world vast amounts of knowledge is stored within many datasets and databases. Often the default format of this data means that the knowledge within is not immediately accessible, but rather has to be mined and extracted. This requires automated tools and they need to be effective and efficient. Association rule mining is one approach to obtaining knowledge stored with datasets / databases which includes frequent patterns and association rules between the items / attributes of a dataset with varying levels of strength. However, this is also association rule mining’s downside; the number of rules that can be found is usually very big. In order to effectively use the association rules (and the knowledge within) the number of rules needs to be kept manageable, thus it is necessary to have a method to reduce the number of association rules. However, we do not want to lose knowledge through this process. Thus the idea of non-redundant association rule mining was born. A second issue with association rule mining is determining which ones are interesting. The standard approach has been to use support and confidence. But they have their limitations. Approaches which use information about the dataset’s structure to measure association rules are limited, but could yield useful association rules if tapped. Finally, while it is important to be able to get interesting association rules from a dataset in a manageable size, it is equally as important to be able to apply them in a practical way, where the knowledge they contain can be taken advantage of. Association rules show items / attributes that appear together frequently. Recommendation systems also look at patterns and items / attributes that occur together frequently in order to make a recommendation to a person. It should therefore be possible to bring the two together. In this thesis we look at these three issues and propose approaches to help. For discovering non-redundant rules we propose enhanced approaches to rule mining in multi-level datasets that will allow hierarchically redundant association rules to be identified and removed, without information loss. When it comes to discovering interesting association rules based on the dataset’s structure we propose three measures for use in multi-level datasets. Lastly, we propose and demonstrate an approach that allows for association rules to be practically and effectively used in a recommender system, while at the same time improving the recommender system’s performance. This especially becomes evident when looking at the user cold-start problem for a recommender system. In fact our proposal helps to solve this serious problem facing recommender systems.

Interestingness Measures for Multi-Level Association Rules

Association rule mining is one technique that is widely used when querying databases, especially those that are transactional, in order to obtain useful associations or correlations among sets of items. Much work has been done focusing on efficiency, effectiveness and redundancy. There has also been a focusing on the quality of rules from single level datasets with many interestingness measures proposed. However, with multi-level datasets now being common there is a lack of interestingness measures developed for multi-level and cross-level rules. Single level measures do not take into account the hierarchy found in a multi-level dataset. This leaves the Support-Confidence approach,which does not consider the hierarchy anyway and has other drawbacks, as one of the few measures available. In this paper we propose two approaches which measure multi-level association rules to help evaluate their interestingness. These measures of diversity and peculiarity can be used to help identify those rules from multi-level datasets that are potentially useful.

Concise representations for association rules in multi-level datasets

Association rule mining has made many advances in the area of knowledge discovery. However, the quality of the discovered association rules is a big concern and has drawn more and more attention recently. One problem with the quality of the discovered association rules is the huge size of the extracted rule set. Often for a dataset, a huge number of rules can be extracted, but many of them can be redundant to other rules and thus useless in practice. Mining non-redundant rules is a promising approach to solve this problem. In this paper, we firstly propose a definition for redundancy; then we propose a concise representation called Reliable basis for representing non-redundant association rules for both exact rules and approximate rules. An important contribution of this paper is that we propose to use the certainty factor as the criteria to measure the strength of the discovered association rules. With the criteria, we can determine the boundary between redundancy and non-redundancy to ensure eliminating as many redundant rules as possible without reducing the inference capacity of and the belief to the remaining extracted non-redundant rules. We prove that the redundancy elimination based on the proposed Reliable basis does not reduce the belief to the extracted rules. We also prove that all association rules can be deduced from the Reliable basis. Therefore the Reliable basis is a lossless representation of association rules. Experimental results show that the proposed Reliable basis can significantly reduce the number of extracted rules.

Interestingness Measures for Multi-Level Association Rules

Association rule mining is one technique that is widely used when querying databases, especially those that are transactional, in order to obtain useful associations or correlations among sets of items. Much work has been done focusing on efficiency, effectiveness and redundancy. There has also been a focusing on the quality of rules from single level datasets with many interestingness measures proposed. However, with multi-level datasets now being common there is a lack of interestingness measures developed for multi-level and cross-level rules. Single level measures do not take into account the hierarchy found in a multi-level dataset. This leaves the Support-Confidence approach, which does not consider the hierarchy anyway and has other drawbacks, as one of the few measures available. In this chapter we propose two approaches which measure multi-level association rules to help evaluate their interestingness by considering the database’s underlying taxonomy. These measures of diversity and peculiarity can be used to help identify those rules from multi-level datasets that are potentially useful.

Mining minimal non-redundant association rules using frequent closed itemsets

The problem of the relevance and the usefulness of extracted association rules is of primary importance because, in the majority of cases, real-life databases lead to several thousands association rules with high confidence and among which are many redundancies. Using the closure of the Galois connection, we define two new bases for association rules which union is a generating set for all valid association rules with support and confidence. These bases are characterized using frequent closed itemsets and their generators; they consist of the non-redundant exact and approximate association rules having minimal antecedents and maximal consequences, i.e. the most relevant association rules. Algorithms for extracting these bases are presented and results of experiments carried out on real-life databases show that the proposed bases are useful, and that their generation is not time consuming.

Non-redundant sequential association rule mining based on closed sequential patterns

In many applications, e.g., bioinformatics, web access traces, system utilisation logs, etc., the data is naturally in the form of sequences. People have taken great interest in analysing the sequential data and finding the inherent characteristics or relationships within the data. Sequential association rule mining is one of the possible methods used to analyse this data. As conventional sequential association rule mining very often generates a huge number of association rules, of which many are redundant, it is desirable to find a solution to get rid of those unnecessary association rules. Because of the complexity and temporal ordered characteristics of sequential data, current research on sequential association rule mining is limited. Although several sequential association rule prediction models using either sequence constraints or temporal constraints have been proposed, none of them considered the redundancy problem in rule mining. The main contribution of this research is to propose a non-redundant association rule mining method based on closed frequent sequences and minimal sequential generators. We also give a definition for the non-redundant sequential rules, which are sequential rules with minimal antecedents but maximal consequents. A new algorithm called CSGM (closed sequential and generator mining) for generating closed sequences and minimal sequential generators is also introduced. A further experiment has been done to compare the performance of generating non-redundant sequential rules and full sequential rules, meanwhile, performance evaluation of our CSGM and other closed sequential pattern mining or generator mining algorithms has also been conducted. We also use generated non-redundant sequential rules for query expansion in order to improve recommendations for infrequently purchased products.

A Multi-layer Perceptron based Non-linear Mixture Model to estimate class abundance from mixed pixels

Sub-pixel classification is essential for the successful description of many land cover (LC) features with spatial resolution less than the size of the image pixels. A commonly used approach for sub-pixel classification is linear mixture models (LMM). Even though, LMM have shown acceptable results, pragmatically, linear mixtures do not exist. A non-linear mixture model, therefore, may better describe the resultant mixture spectra for endmember (pure pixel) distribution. In this paper, we propose a new methodology for inferring LC fractions by a process called automatic linear-nonlinear mixture model (AL-NLMM). AL-NLMM is a three step process where the endmembers are first derived from an automated algorithm. These endmembers are used by the LMM in the second step that provides abundance estimation in a linear fashion. Finally, the abundance values along with the training samples representing the actual proportions are fed to multi-layer perceptron (MLP) architecture as input to train the neurons which further refines the abundance estimates to account for the non-linear nature of the mixing classes of interest. AL-NLMM is validated on computer simulated hyperspectral data of 200 bands. Validation of the output showed overall RMSE of 0.0089±0.0022 with LMM and 0.0030±0.0001 with the MLP based AL-NLMM, when compared to actual class proportions indicating that individual class abundances obtained from AL-NLMM are very close to the real observations.

Using Association Rules to Solve the Cold-Start Problem in Recommender Systems

Recommender systems are widely used online to help users find other products, items etc that they may be interested in based on what is known about that user in their profile. Often however user profiles may be short on information and thus it is difficult for a recommender system to make quality recommendations. This problem is known as the cold-start problem. Here we investigate using association rules as a source of information to expand a user profile and thus avoid this problem. Our experiments show that it is possible to use association rules to noticeably improve the performance of a recommender system under the cold-start situation. Furthermore, we also show that the improvement in performance obtained can be achieved while using non-redundant rule sets. This shows that non-redundant rules do not cause a loss of information and are just as informative as a set of association rules that contain redundancy.

A Reliable Basis for Approximate Association Rules

For most of the work done in developing association rule mining, the primary focus has been on the efficiency of the approach and to a lesser extent the quality of the derived rules has been emphasized. Often for a dataset, a huge number of rules can be derived, but many of them can be redundant to other rules and thus are useless in practice. The extremely large number of rules makes it difficult for the end users to comprehend and therefore effectively use the discovered rules and thus significantly reduces the effectiveness of rule mining algorithms. If the extracted knowledge can’t be effectively used in solving real world problems, the effort of extracting the knowledge is worth little. This is a serious problem but not yet solved satisfactorily. In this paper, we propose a concise representation called Reliable Approximate basis for representing non-redundant approximate association rules. We prove that the redundancy elimination based on the proposed basis does not reduce the belief to the extracted rules. We also prove that all approximate association rules can be deduced from the Reliable Approximate basis. Therefore the basis is a lossless representation of approximate association rules.

Reliable representations for association rules

Association rule mining has contributed to many advances in the area of knowledge discovery. However, the quality of the discovered association rules is a big concern and has drawn more and more attention recently. One problem with the quality of the discovered association rules is the huge size of the extracted rule set. Often for a dataset, a huge number of rules can be extracted, but many of them can be redundant to other rules and thus useless in practice. Mining non-redundant rules is a promising approach to solve this problem. In this paper, we first propose a definition for redundancy, then propose a concise representation, called a Reliable basis, for representing non-redundant association rules. The Reliable basis contains a set of non-redundant rules which are derived using frequent closed itemsets and their generators instead of using frequent itemsets that are usually used by traditional association rule mining approaches. An important contribution of this paper is that we propose to use the certainty factor as the criterion to measure the strength of the discovered association rules. Using this criterion, we can ensure the elimination of as many redundant rules as possible without reducing the inference capacity of the remaining extracted non-redundant rules. We prove that the redundancy elimination, based on the proposed Reliable basis, does not reduce the strength of belief in the extracted rules. We also prove that all association rules, their supports and confidences, can be retrieved from the Reliable basis without accessing the dataset. Therefore the Reliable basis is a lossless representation of association rules. Experimental results show that the proposed Reliable basis can significantly reduce the number of extracted rules. We also conduct experiments on the application of association rules to the area of product recommendation. The experimental results show that the non-redundant association rules extracted using the proposed method retain the same inference capacity as the entire rule set. This result indicates that using non-redundant rules only is sufficient to solve real problems needless using the entire rule set.

Defining Interestigness for Association Rules

Interestingness in Association Rules has been a major topic of research in the past decade. The reason is that the strength of association rules, i.e. its ability to discover ALL patterns given some thresholds on support and confidence, is also its weakness. Indeed, a typical association rules analysis on real data often results in hundreds or thousands of patterns creating a data mining problem of the second order. In other words, it is not straightforward to determine which of those rules are interesting for the end-user. This paper provides an overview of some existing measures of interestingness and we will comment on their properties. In general, interestingness measures can be divided into objective and subjective measures. Objective measures tend to express interestingness by means of statistical or mathematical criteria, whereas subjective measures of interestingness aim at capturing more practical criteria that should be taken into account, such as unexpectedness or actionability of rules. This paper only focusses on objective measures of interestingness.

Effective hybrid recommendation combining users-searches correlations using tensors

Most recommendation methods employ item-item similarity measures or use ratings data to generate recommendations. These methods use traditional two dimensional models to find inter relationships between alike users and products. This paper proposes a novel recommendation method using the multi-dimensional model, tensor, to group similar users based on common search behaviour, and then finding associations within such groups for making effective inter group recommendations. Web log data is multi-dimensional data. Unlike vector based methods, tensors have the ability to highly correlate and find latent relationships between such similar instances, consisting of users and searches. Non redundant rules from such associations of user-searches are then used for making recommendations to the users.

Efficient search for statistically significant dependency rules in binary data

Analyzing statistical dependencies is a fundamental problem in all empirical science. Dependencies help us understand causes and effects, create new scientific theories, and invent cures to problems. Nowadays, large amounts of data is available, but efficient computational tools for analyzing the data are missing. In this research, we develop efficient algorithms for a commonly occurring search problem - searching for the statistically most significant dependency rules in binary data. We consider dependency rules of the form X->A or X->not A, where X is a set of positive-valued attributes and A is a single attribute. Such rules describe which factors either increase or decrease the probability of the consequent A. A classical example are genetic and environmental factors, which can either cause or prevent a disease. The emphasis in this research is that the discovered dependencies should be genuine - i.e. they should also hold in future data. This is an important distinction from the traditional association rules, which - in spite of their name and a similar appearance to dependency rules - do not necessarily represent statistical dependencies at all or represent only spurious connections, which occur by chance. Therefore, the principal objective is to search for the rules with statistical significance measures. Another important objective is to search for only non-redundant rules, which express the real causes of dependence, without any occasional extra factors. The extra factors do not add any new information on the dependence, but can only blur it and make it less accurate in future data. The problem is computationally very demanding, because the number of all possible rules increases exponentially with the number of attributes. In addition, neither the statistical dependency nor the statistical significance are monotonic properties, which means that the traditional pruning techniques do not work. As a solution, we first derive the mathematical basis for pruning the search space with any well-behaving statistical significance measures. The mathematical theory is complemented by a new algorithmic invention, which enables an efficient search without any heuristic restrictions. The resulting algorithm can be used to search for both positive and negative dependencies with any commonly used statistical measures, like Fisher's exact test, the chi-squared measure, mutual information, and z scores. According to our experiments, the algorithm is well-scalable, especially with Fisher's exact test. It can easily handle even the densest data sets with 10000-20000 attributes. Still, the results are globally optimal, which is a remarkable improvement over the existing solutions. In practice, this means that the user does not have to worry whether the dependencies hold in future data or if the data still contains better, but undiscovered dependencies.