Heuristic Mining Research Articles

Modified Time-Based Heuristics Miner for Parallel Business Processes

Process Mining, or Process Discovery, is a method for modeling the workflow of a business process from event logs. Business process models contain sequential and parallel traces. In this paper, a modification of the frequently used process-mining algorithm Heuristics Miner is proposed. The proposed algorithm is called Modified Time-based Heuristics Miner because it considers not only the sequence of activities but also the time-based information from the event log. It can effectively distinguish parallel (AND), single choice (XOR) and conditional (OR) patterns; the latter cannot be discovered by the original Heuristics Miner. The threshold intervals are determined on the basis of the average dependency measure in the dependency graph. The experimental results show that the proposed algorithm is able to discover concurrent business processes formed by parallel (AND) and conditional (OR) patterns, whereas the existing Heuristics Miner algorithm can only discover concurrent business processes formed by parallel (AND) patterns. This paper also provides an evaluation of validity an fitness of the discovered process model.

Adaptive Process Simplification and Conformance Checking in Event Logs

In this paper, we used two classes of process mining techniques (i.e., Discovery and Conformance Analysis) in order to discover models and organizational structures related to the handling of proceedings’ peer reviews in an international conference in Thailand. We monitored the deviations, comparing the observed events (real-life data) with the predefined models, as well. Knowing that ProM plugins receive the input logs only in MXML (Mining eXtensible Markup Language) and XES (eXtensible Event Stream) formats, we chose ProMimport as a framework to extract MXML log from an event log consisted of 87 cases (papers) and 3267 events. Alpha (α) Algorithm, Heuristic, Fuzzy and Social Network mining techniques (from Discovery class) were applied to automatically construct the proceedings’ review models, primarily without any priori model. Though the Heuristic Miner closely followed the Alpha algorithm, the technique had the privilege to derive XOR and AND connectors from dependency relations of the event log. Next, the actual process behavior was projected onto fuzzy models. The result was an animation movie coming up with a better understanding of what has occurred in reality. Also, using Social Network Miner technique we could analyze the organizational perspective of the peer review process in terms of three metrics, namely as: (a) Handover of Work, (b) Working Together, and (c) Similar Tasks. Alternatively, having a priori model for review of the proceedings, we used LTL Checker and Performance Analysis techniques (from Conformance Checker class) to identify discrepancies between the log and the pre-defined model. After applying the LTL checking approach, the deviations were detected — leading to enrich the real model. In addition, Performance Analysis technique made us capable of projecting the bottlenecks all through the peer review system. In general, one of the main benefits of the techniques used in this paper is that information is objectively compiled. To say simple, we gathered valuable information about what actually was happening according to the review process of the papers and existing bottlenecks, and not what we just thought or expected to see happening in the event log. Considering the results of the study, conference committee chairs can better evaluate the performance of the involving reviewers (as well as team members) within the assigned tasks. This will improve the performance, efficiency and effectiveness of the handling of reviews for prospective academic/educational conferences.

Improving the quality of the Heuristics Miner in ProM 6.2

Considering the presence of large amounts of data in organizations today, the need to transform this data into useful information and subsequently into knowledge, increasingly gains attention. Process discovery is a technique to automatically discover process models from data in event logs. Since process discovery is gaining attention among researchers as well as practitioners, the quality of the resulting process model must be assured. In this paper, the quality of the frequently used Heuristics Miner is improved as anomalies were found concerning the validity and completeness of the resulting process model. For this purpose, a new artifact called the Updated Heuristics Miner was constructed containing alterations to the tool and to the algorithm itself. Evaluations of this artifact resulted in the conclusion that the Updated Heuristics Miner indeed demonstrates higher validity and completeness. This study contributes to the body of knowledge first by improving the quality of the an often used research instrument and second by stating that there is a need for a systematic developing and evaluation method for process discovery techniques.

A Heuristic Mining Algorithm Using Web Hyperlink Structure

Web mining aims to discover useful information or knowledge from the Web hyperlink structure, page content and usage log. Based on the primary kind of data used in the mining process, Web mining tasks are categorized into three main types: Web structure mining, Web content mining and Web usage mining. Following is what they do on Web Data Mining. This paper proposed a heuristic mining algorithm.

Analyzing miRNA co-expression networks to explore TF-miRNA regulation

BackgroundCurrent microRNA (miRNA) research in progress has engendered rapid accumulation of expression data evolving from microarray experiments. Such experiments are generally performed over different tissues belonging to a specific species of metazoan. For disease diagnosis, microarray probes are also prepared with tissues taken from similar organs of different candidates of an organism. Expression data of miRNAs are frequently mapped to co-expression networks to study the functions of miRNAs, their regulation on genes and to explore the complex regulatory network that might exist between Transcription Factors (TFs), genes and miRNAs. These directions of research relating miRNAs are still not fully explored, and therefore, construction of reliable and compatible methods for mining miRNA co-expression networks has become an emerging area. This paper introduces a novel method for mining the miRNA co-expression networks in order to obtain co-expressed miRNAs under the hypothesis that these might be regulated by common TFs.ResultsThree co-expression networks, configured from one patient-specific, one tissue-specific and a stem cell-based miRNA expression data, are studied for analyzing the proposed methodology. A novel compactness measure is introduced. The results establish the statistical significance of the sets of miRNAs evolved and the efficacy of the self-pruning phase employed by the proposed method. All these datasets yield similar network patterns and produce coherent groups of miRNAs. The existence of common TFs, regulating these groups of miRNAs, is empirically tested. The results found are very promising. A novel visual validation method is also proposed that reflects the homogeneity as well as statistical properties of the grouped miRNAs. This visual validation method provides a promising and statistically significant graphical tool for expression analysis.ConclusionA heuristic mining methodology that resembles a clustering motivation is proposed in this paper. However, there remains a basic difference between the mining method and a clustering approach. The heuristic approach can produce priority modules (PM) from an miRNA co-expression network, by employing a self-pruning phase, which are analyzed for statistical and biological significance. The mining algorithm minimizes the space/time complexity of the analysis, and also handles noise in the data. In addition, the mining method reveals promising results in the unsupervised analysis of TF-miRNA regulation.

Modeling and analysis for multi-period, multi-product and multi-resource production scheduling

This study presents a framework for solving the multi-period, multi-product and multi-resource production-scheduling (M3PS) problem. Practically, the main concern for an M3PS problem is how to satisfy two management policies: (1) each product is manufactured in a continuous manner so that once the product is on a production line, it will complete its production procedure without interruption, and (2) the number of the product's types is limited during one period. By defining the decision variables and taking into account the machine's capacity and the customers' demand, a mixed integer programming (MIP) Model is formulated. To solve this MIP problem, a two-phase approach is proposed. In phase 1, the search space of the MIP Model is transformed into a preliminary pattern by a heuristic mining algorithm so that a hyper assignment problem can be formed as a reference model to be solved. In phase 2, a stochastic global optimization procedure that incorporates a genetic algorithm with neighborhood search techniques is designed to obtain the optimal solution. A numerical experiment is presented with an illustration, and it shows that the proposed model is adequate to cope with complicate scheduling problems.