Search

The proposed work aims to solve data sparsity problem in the recommendation system. It handles two-level pre-processing techniques to reduce the data size at the item level. Additional resources like items genre, tag, and time are added to learn and analyse the behaviour of the user preferences in-depth. The advantage of the proposed method is to recommend the item, based on user interest pattern and avoid recommending the outdated items. User information are grouped based on similar item genre and tag feature. This effectively handle overlapping conditions that exist on item's genre, as it has more than one genre at initial level. Further, based on time, it analyses the user non-static interest. Overall it reduces the dimensions which is an initial way to prepare data, to analyse hidden pattern. To enhance the performance, the proposed method utilized Apache's spark Mllib FP-Growth and association rule mining approach in a distributed environment. To reduce the computation cost of constructing tree in FP-Growth, the candidate data set is stored in matrix form. The experiments were conducted using MovieLens data set. The observed results shows that the proposed method achieves 4% increase in accuracy when compared to earlier methods. [ABSTRACT FROM AUTHOR]

Meteorological disasters pose a significant challenge to megacities, and quantitatively identifying the influence of meteorological conditions on disasters is an urgent issue. Compared to other traditional datasets, disaster reporting data offers significant informational value for urban meteorological risk response and assessment due to its high credibility and timeliness. This study is based on meteorological disaster reporting data for Shanghai over the past 10 years and focuses on the identification and threshold analysis of meteorological disaster risk factors for disaster-affected entities, offering a novel perspective on analyzing disaster-causing risk under the combined influence of meteorological conditions. The findings suggest that disaster reporting data can be used for meteorological risk assessment. There were noticeable differences in the disaster thresholds for affected entities in central urban areas (city center) versus suburban areas (suburbs). The combined effects of strong winds and heavy rain were often significant contributors to the occurrence of disasters. Furthermore, several distinct disaster characteristics were identified through frequent pattern (FP)-growth mining. The results of this study provide technical support for early warning and disaster management under specific weather conditions by constructing a quantitative framework for understanding the impact of hazardous weather. • Disaster reporting data can provide valuable information for analyzing disaster impact characteristics. • A 10-year database composed of 51,409 disaster reporting data is used for analysis. • FP-growth algorithm extracted frequent patterns between meteorological conditions and disaster onset on affected entities. [ABSTRACT FROM AUTHOR]

Abstract In Electric Management Information System (MIS), there are some users who do not comply with all operation/behavior regulations and make the similar mistakes continuously even though they are not on purpose. These behaviors are a huge threat to the system security. In this paper, we propose a method to detect these regular behaviors with association rules mining algorithm FP-Growth. First, the user log is separated into operation sets each of which contains user operation in a continuous period. Then we divide the operation sets of all users into two catalogs: normal and abnormal based on if a security problem has happened around the corresponding period of operation set. Next, we apply the FP-Growth algorithm in both normal and abnormal operation sets to generate the frequent patterns. Finally, the abnormal pattern is compared with normal ones to determine the regular behaviors that may be dangerous to the system. We test the proposed algorithm in the user log files generated from a simulated electric management information system. The experiment results indicate the proposed method can effectively detect the regular user behavior that could cause the system security problems. [ABSTRACT FROM AUTHOR]

• To solve the problem how to activate rules that have high confidence but low support, the information gain (IG) concept is introduced in the model building process. • The proposed model consists of a set of decision rules that formally represent behavioral scripts, for predicting individuals' transportation mode choice. • A 10-fold cross validation test using a grid search parameter optimization method is conducted to validate the proposed approach. Predicting transportation mode choice is a classic challenge of travel behavior research. Over the years, different theoretical concepts and modeling approaches have been applied. This paper elaborates the application of class association rules (CARs) and examines their predictive performance using data extracted from the 2015 National Dutch Travel Survey. To solve the problem how to activate rules that have high confidence but low support, the information gain (IG) concept is introduced in the model building process. The modeling process in this study first involves extracting frequent items from the data using the FP-Growth algorithm and deriving CARs from these frequent items. Next, the IG statistic is used to construct a novel model (named CARIG), which consists of a set of decision rules that formally represent behavioral scripts, for predicting individuals' transportation mode choice. The performance of CARIG is compared with the performance of conventional class-based association rules (CBA), decision trees (DT), a convolutional neural network (CNN) and a logistic regression (LR) model. In addition, a 10-fold cross validation test using a grid search parameter optimization method is conducted to validate the proposed approach. The results show that the proposed method is promising in predicting transportation mode choices observed in the national travel survey data. [ABSTRACT FROM AUTHOR]

Automation in pharmacies has achieved innovative levels of effectiveness and savings. In the present day, automated pharmacies are facing extremely large demands of prescription orders specifically at the central fill pharmacies that distribute drugs to retail pharmacies. As a result, improvements are necessary to the Robotic Prescription Dispensing System (RPDS) and RPDS planogram to increase the throughput of prescriptions. RPDS planogram defines where to allocate the dispensers inside the robotic unit and how to distribute them among the multiple robotic units. This research utilizes the pharmacy prescriptions database to extract useful knowledge to improve different strategies in pharmacy automation by using a data mining approach. In this study, a data mining tool is proposed to enhance pharmacy automation. Frequent Pattern Growth (FP-growth) approach, which is one of the algorithms of Association Rule Mining (ARM), is applied to an actual prescriptions database of a central fill pharmacy to study the associations within the prescribed drug regime. The FP-growth application in a prescriptions database is novel; thus, FP-growth is tested on both sequential mode, and parallel mode by using a distributed platform Hadoop and MapReduce paradigm. Two types of association rules are extracted: 1) associations among different drugs that involve their different dosage strengths and manufacturers; and 2) associations that include only information about different drug generic and brand names. The importance of the extracted association rules is evaluated by the use of different measures, including the support, confidence, lift and conviction. The discovered rules disclose strong associations among the purchased drugs that improve the allocation and distribution of dispensers among the robotic units, in addition to enhancements in other pharmacy managerial strategies. [ABSTRACT FROM AUTHOR]

Identifying related offences in a criminal investigation is an important goal for crime analysts. This can deliver evidence that can assist in apprehension of suspects and better attribution of past crimes. The use of pattern based approaches has the potential to assist crime experts in discovering new patterns of criminal activity. Hence, research in this area continues. This paper revisits frequent pattern growth models for crime pattern mining. Frequent pattern (FP) based approaches, such as the FP-Growth model, have been identified to be more effective than techniques proposed in the past, such as Apriori. Therefore, this research proposes a descriptive statistical approach, based on a quartile (floor-ceil) function, for the minimum support threshold (MST) choice selection, which is a major decision step in the pruning phase of the Traditional FP-Growth (TFPG) model. Our revised frequent pattern growth (RFPG) model further proposes a Pattern-pattern ( Pp ) paradigm to identify tuples of subtle crime pattern(s) sequences or recurring trends in criminal activity. We present empirical results in order to guide intended audience about future decisions or research regarding this model. Results indicate that RFPG is more promising than TFPG and will always ensure the utilisation of a reasonable percentage of the crime dataset, in order to produce more reliable and sufficiently informative patterns or trends. [ABSTRACT FROM AUTHOR]

Abstract: FP-Growth is a recursive mining algorithm for frequent patterns. With each recursion, a new FP-tree and the corresponding header table are built. This paper proposes another approach to improve the mining efficiency: only one FP-tree is constructed, and instead of building new trees in every recursion, builds new header tables; frequent patterns are generated along the way, so we call this algorithm “header table recursion (HTR)” method. Tests show that it will get better performance for relative small datasets because the compress of the FP-tree has no significant impact on the performance yet inflicts extra time cost in tree creation operations. However on larger datasets, the traversal of the original FP-tree will become more time-consuming and its overall performance needs further testing. [Copyright &y& Elsevier]