Sequential Database Research Articles

A contribution to the discovery of multidimensional patterns in healthcare trajectories

Sequential pattern mining is aimed at extracting correlations among temporal data. Many different methods were proposed to either enumerate sequences of set valued data (i.e., itemsets) or sequences containing dimensional items. However, in real-world scenarios, data sequences are described as combination of both multidimensional items and itemsets. These heterogeneous descriptions cannot be handled by traditional approaches. In this paper we propose a new approach called MMISP (Mining Multidimensional Itemset Sequential Patterns) to extract patterns from complex sequential database including both multidimensional items and itemsets. The novelties of the proposal lies in: (i) the way in which the data are efficiently compressed; (ii) the ability to reuse and adopt sequential pattern mining algorithms and (iii) the extraction of new kind of patterns. We introduce a case-study on real-world data from a regional healthcare system and we point out the usefulness of the extracted patterns. Additional experiments on synthetic data highlights the efficiency and scalability of the approach MMISP.

Predictors of Career Choice in Academic Medicine Among Orthopaedic Surgery Residents

In 2013, 1038 applicants, 833 of whom were U.S. graduates, applied for 693 orthopaedic residency spots. Orthopaedic residency program directors thoroughly screen applicant data, selecting those applicants who will be most successful during the residency. Although the most important consideration for program directors is choosing residents who will obtain competence and will pass the orthopaedic board examinations, many applicants are chosen also for their ability to contribute to the field through leadership, research, and teaching. Previous studies have analyzed attitudes of orthopaedic residents toward research and university-based careers1. Primary authorship of a publication and past research experience were associated with a greater interest in future research1. However, to the best of our knowledge, no study has been done to identify predictors of orthopaedic residents who actually obtain faculty positions in a university-based setting. We hypothesized that predictive factors exist for orthopaedic surgery residents who choose a university-based career. This study looked at both academic and socioeconomic criteria to identify factors that predict residents becoming assistant professors, associate professors, full professors, program directors, or program chairs. This study was approved by the institutional review board. It was performed in compliance with the Health Insurance Portability and Accountability Act regulations. ### Identifying Career Outcomes Among Orthopaedic Residency Graduates The orthopaedic surgery department at our hospital compiled a complete and sequential database of sixty residents who graduated from 2001 to 2010 and their career choices. No residents were excluded. If career choice data were not available for a former resident, a Google search was performed to find out the highest academic rank achieved by the graduate. Former residents were contacted by e-mail if they could not be found on a Google search. The careers of orthopaedic surgeons were analyzed and were separated into a university or community-based career choice. University-based orthopaedic surgeons had a full-time faculty appointment …

Medical Product Safety Surveillance

Large linked database networks, like the US Food and Drug Administration's Sentinel System, are being built for medical product surveillance. One use of these networks is for "near real-time" sequential database surveillance of prespecified medical product-adverse event pairs, which may result in a "safety signal" when a statistical excess risk is detected. Sequential database surveillance requires the investigator to manage surveillance in both information time (ie, how sample size accrues) and calendar time. Calendar time is important because people external to the surveillance population may be affected by the speed with which a safety signal is detected or ruled out. Optimal design and analysis aspects of sequential database surveillance are not well developed, but are gaining in importance as observational database networks grow. Using information time concepts, we show how to calculate sample sizes when performing sequential database surveillance, illustrating the relationships between statistical power, the time to detect a signal, and the maximum sample size for various true effect sizes. Then, using a vaccine example, we demonstrate a four-step planning process that allows investigators to translate information time into calendar time. Given the calendar time for surveillance, the process focuses on choosing observational database configurations consistent with the investigator's preferences for timeliness and statistical power. Although the planning process emphasizes sample size considerations, the influence of secondary database attributes such as delay times, measurement error, and cost are also discussed. Appropriate planning allows the most efficient use of public health dollars dedicated to medical product surveillance efforts.

A data mining approach to discovering reliable sequential patterns

Sequential pattern mining is a data mining method for obtaining frequent sequential patterns in a sequential database. Conventional sequence data mining methods could be divided into two categories: Apriori-like methods and pattern growth methods. In a sequential pattern, probability of time between two adjacent events could provide valuable information for decision-makers. As far as we know, there has been no methodology developed to extract this probability in the sequential pattern mining process. We extend the PrefixSpan algorithm and propose a new sequential pattern mining approach: P-PrefixSpan. Besides minimum support-count constraint, this approach imposes minimum time-probability constraint, so that fewer but more reliable patterns will be obtained. P-PrefixSpan is compared with PrefixSpan in terms of number of patterns obtained and execution efficiency. Our experimental results show that P-PrefixSpan is an efficient and scalable method for sequential pattern mining.

Incremental Mining of Sequential Patterns Using Weights

Real life sequential databases are usually not static. They tend to grow incrementally. So after every update a frequent pattern may no longer remains frequent while some infrequent patterns may appear as frequent in updated database. It is not a good idea to mine sequential database from scratch every time as the update occurs. It would be better if we can use the knowledge of already mined sequential patterns to find the complete set of sequential patterns for updated database. An incremental mining algorithm does the same thing. The main goal of an incremental mining algorithm is to reduce the time taken to find out the frequent patterns significantly i.e. it should mine the set of frequent patterns in significantly less time than a non-incremental mining algorithm. In this work we have approached using weight constraints, in time and space, of an idea of already existing algorithm called WSM.

Using high‐dimensional propensity scores to automate confounding control in a distributed medical product safety surveillance system

Distributed medical product safety monitoring systems such as the Sentinel System, to be developed as a part of Food and Drug Administration's Sentinel Initiative, will require automation of large parts of the safety evaluation process to achieve the necessary speed and scale at reasonable cost without sacrificing validity. Although certain functions will require investigator intervention, confounding control is one area that can largely be automated. The high-dimensional propensity score (hd-PS) algorithm is one option for automated confounding control in longitudinal healthcare databases. In this article, we discuss the use of hd-PS for automating confounding control in sequential database cohort studies, as applied to safety monitoring systems. In particular, we discuss the robustness of the covariate selection process, the potential for over- or under-selection of variables including the possibilities of M-bias and Z-bias, the computation requirements, the practical considerations in a federated database network, and the cases where automated confounding adjustment may not function optimally. We also outline recent improvements to the algorithm and show how the algorithm has performed in several published studies. We conclude that despite certain limitations, hd-PS offers substantial advantages over non-automated alternatives in active product safety monitoring systems.

Impact of exposure accrual on sequential postmarket evaluations: a simulation study

We frame the challenges in conducting sequential database surveillance (SDS) analyses and use simulation techniques to illustrate one of these aspects: the accrual of exposed person-time. We discuss the role of SDS analyses in the US Food and Drug Administration's planned Sentinel System and suggest the outline of a decision-analytic framework that might aid regulatory authorities in planning postmarket surveillance. We trace exposure accrual using US Medicaid data for new molecular entities approved in 2005 and 2006 and simulate risk estimates that are observable using SDS by varying the event rate of the health outcome of interest (HOI) in the control group. Sequential database surveillance may considerably narrow risk estimates for most products within 2 years when the underlying control group HOI occurs commonly (i.e., at least 1 HOI/100 person-years). Exposure accrual may be insufficient for detecting relative risks with infrequent, rare, or very rare HOIs (i.e., less frequent than 1 event/1000 person-years). However, if SDS is used to quickly detect or rule out an incremental risk difference, then a majority of products are viable SDS candidates. As a surveillance tool, SDS performs differently depending on whether its intended regulatory target is to narrow relative risk estimates or to confirm/rule out absolute levels of excess risk. These desired endpoints, in addition to the number of exposures available for analysis and the underlying HOI rates, influence the number of products that would be viable candidates for SDS.

Binary Coded Web Access Pattern Tree in Education Domain

Web Access Pattern (WAP), which is the sequence of accesses pursued by users frequently, is a kind of interesting and useful knowledge in practice. Sequential Pattern mining is the process of applying data mining techniques to a sequential database for the purposes of discovering the correlation relationships that exist among an ordered list of events. WAP tree mining is a sequential pattern mining technique for web log access sequences, which first stores the original web access sequence database on a prefix tree. WAP-tree algorithm then, mines the frequent sequences from the WAP-tree by recursively re-constructing intermediate trees. In this paper, we propose efficient sequential pattern techniques called BC-WAP (Binary Coded WAP). The proposed algorithm uses Kongu Arts and Science College web logs for sequential pattern mining. It eliminates recursively reconstructing intermediate WAP trees during the mining by assigning the binary codes to each node in the WAP tree. The results of the experiments show the efficiency of the improved algorithm.

On Mining Instance-Centric Classification Rules

Many studies have shown that rule-based classifiers perform well in classifying categorical and sparse high-dimensional databases. However, a fundamental limitation with many rule-based classifiers is that they find the rules by employing various heuristic methods to prune the search space and select the rules based on the sequential database covering paradigm. As a result, the final set of rules that they use may not be the globally best rules for some instances in the training database. To make matters worse, these algorithms fail to fully exploit some more effective search space pruning methods in order to scale to large databases. In this paper, we present a new classifier, HARMONY, which directly mines the final set of classification rules. HARMONY uses an instance-centric rule-generation approach and it can assure that, for each training instance, one of the highest-confidence rules covering this instance is included in the final rule set, which helps in improving the overall accuracy of the classifier. By introducing several novel search strategies and pruning methods into the rule discovery process, HARMONY also has high efficiency and good scalability. Our thorough performance study with some large text and categorical databases has shown that HARMONY outperforms many well-known classifiers in terms of both accuracy and computational efficiency and scales well with regard to the database size.

Finding Frequent Patterns Using Length-Decreasing Support Constraints

Finding prevalent patterns in large amount of data has been one of the major problems in the area of data mining. Particularly, the problem of finding frequent itemset or sequential patterns in very large databases has been studied extensively over the years, and a variety of algorithms have been developed for each problem. The key feature in most of these algorithms is that they use a constant support constraint to control the inherently exponential complexity of these two problems. In general, patterns that contain only a few items will tend to be interesting if they have a high support, whereas long patterns can still be interesting even if their support is relatively small. Ideally, we want to find all the frequent patterns whose support decreases as a function of their length without having to find many uninteresting infrequent short patterns. Developing such algorithms is particularly challenging because the downward closure property of the constant support constraint cannot be used to prune short infrequent patterns. In this paper we present two algorithms, LPMiner and SLPMiner. Given a length-decreasing support constraint, LPMiner finds all the frequent itemset patterns from an itemset database, and SLPMiner finds all the frequent sequential patterns from a sequential database. Each of these two algorithms combines a well-studied efficient algorithm for constant-support-based pattern discovery with three effective database pruning methods that dramatically reduce the runtime. Our experimental evaluations show that both LPMiner and SLPMiner, by effectively exploiting the length-decreasing support constraint, are up to two orders of magnitude faster, and their runtime increases gradually as the average length of the input patterns increases.

Mining Web Log Sequential Patterns with Position Coded Pre-Order Linked WAP-Tree

Sequential mining is the process of applying data mining techniques to a sequential database for the purposes of discovering the correlation relationships that exist among an ordered list of events. An important application of sequential mining techniques is web usage mining, for mining web log accesses, where the sequences of web page accesses made by different web users over a period of time, through a server, are recorded. Web access pattern tree (WAP-tree) mining is a sequential pattern mining technique for web log access sequences, which first stores the original web access sequence database on a prefix tree, similar to the frequent pattern tree (FP-tree) for storing non-sequential data. WAP-tree algorithm then, mines the frequent sequences from the WAP-tree by recursively re-constructing intermediate trees, starting with suffix sequences and ending with prefix sequences. This paper proposes a more efficient approach for using the WAP-tree to mine frequent sequences, which totally eliminates the need to engage in numerous re-construction of intermediate WAP-trees during mining. The proposed algorithm builds the frequent header node links of the original WAP-tree in a pre-order fashion and uses the position code of each node to identify the ancestor/descendant relationships between nodes of the tree. It then, finds each frequent sequential pattern, through progressive prefix sequence search, starting with its first prefix subsequence event. Experiments show huge performance gain over the WAP-tree technique.

Discovering Patterns from Large and Dynamic Sequential Data

Most daily and scientific data are sequential in nature. Discovering important patterns from such data can benefit the user and scientist by predicting coming activities, interpreting recurring phenomena, extracting outstanding similarities and differences for close attention, compressing data, and detecting intrusion. We consider the following incremental discovery problem for large and dynamic sequential data. Suppose that patterns were previously discovered and materialized. An update is made to the sequential database. An incremental discovery will take advantage of discovered patterns and compute only the change by accessing the affected part of the database and data structures. In addition to patterns, the statistics and position information of patterns need to be updated to allow further analysis and processing on patterns. We present an efficient algorithm for the incremental discovery problem. The algorithm is applied to sequential data that honors several sequential patterns modeling weather changes in Singapore. The algorithm finds what it is supposed to find. Experiments show that for small updates and large databases, the incremental discovery algorithm runs in time independent of the data size.

Thrifty execution of task pipelines

A sequence of tasks that must be performed on a sequential database can be scheduled in various ways. Schedules will differ with respect to the number of accesses made to peripheral storage devices and the amount of memory space consumed by buffers. Buffer requirements are discussed for task schedules that avoid accesses to peripherals storing the sequential database. The relationship between certain thrifty scheduling policies and loop jamming, a standard code optimization technique, is also identified. Application to UNIX pipelines and to file processing is discussed.

Individual Difference Effects in Human-Computer Interaction

One of the most common assumptions in design of Human-Computer interfaces is that there is an advantage to matching the psychological (cognitive) characteristics of the user to the interface features of the computer. However, except in the area of training and experience there have been few clear demonstrations that underlying cognitive characteristics of the user significantly interact with interface features. This paper will describe an experiment in which the effects of individual differences in user's information (cognitive) style (specifically field independence/dependence and breadth of categorization) upon interaction with a complex computer program were examined. The program was a realistically difficult personal finance management system which required database searching and problem solving activities. The program was rich enough to allow alternative interaction methods. The experiment comprised three parts. In the first phase, subjects learned how to use the computer system. In the second phase the subjects entered information into a personal finance database and in the third phase the subjects used their personal database to locate and correct discrepancies between a simulated bank statement and financial records. Prior to these experimental tasks subjects information-cognitive styles were evaluated using standardized tests of field independence/field dependence and breadth of categorization. The results indicate that users with different underlying information styles, at similar levels of learning, have different preferred methods of interacting with the computer, and, that different database search methods are preferred by users of different cognitive styles. For example, field independent subjects preferred key-word multidimensional, non sequential database search methods while field dependent subjects used a method in which the database was searched in a linear, sequential mode. This research is sponsored by US Army Research Institute for the Behavioral and Social Sciences

Marine geophysics: Database management and supportive graphics

The National Geophysical Data Center (NGDC), a national and international repository for marine geophysical data, has developed a unique database-management system known as GEODAS (Geophysical Data System). GEODAS is a user-friendly computer system that integrates an index sequential database with an interactive (random access) inventory and data-documentation file. The system is supported by several graphic peripherals, including cathode-ray tube (CRT) terminals, computer-output-microfilm (COM), and plotters. GEODAS incorporates computer software routines that (a) process underway marine geophysical data (i.e. bathymetry, magnetics, and gravity) into a common-format database, (b) provide both graphical and textual inventory information, and (c) retrieve data using multiparameter select criteria. The GEODAS inventory now is used as a remote access system by research institutions throughout the United States. In the future, we hope to provide remote access to the data file as well and to expand the inventory portion to act as an international referral system.