一种基于频繁序列树的增量式序列模式挖掘算法

针对目前现有的增量式序列模式挖掘算法没有充分利用先前的挖掘结果,当数据库更新时,需要对数据库进行重复挖掘的问题。本文提出一种基于频繁序列树的增量式序列模式挖掘算法(ISFST),ISFST采用频繁序列树作为序列存储结构,当数据库发生变化时,ISFST算法分两种情况对频繁序列树进行更新操作,通过遍历频繁序列树得到满足最小支持度的所有序列模式。实验结果表明,ISFST算法在时间性能上优于PrefixSpan算法和IncSpan算法。     

基于序列树的增量式序列模式更新算法

在序列数据库更新时,现有的增量式序列模式挖掘算法只提到序列的插入操作和序列的扩展操作两种情况,没有针对序列删除操作。提出了一种基于序列树的增量式序列模式更新算法（ISPST）。当数据库更新时,ISPST算法只需要对与删除序列有关的序列构造投影数据库,实现对序列树的更新操作,通过深度优先遍历序列树得到更新后数据库中的所有序列模式。实验结果表明,当支持度发生变化时,ISPST算法在时间性能上优于PrefixSpan算法和IncSpan算法。     

一种高效的增量式序列模式挖掘算法

现有的增量式挖掘算法在支持度发生变化时,需要对序列数据库进行重复挖掘,为减少由此产生的时空消耗,提出一种高效的增量式序列模式挖掘算法。算法采用频繁序列树作为序列存储结构,当序列数据库和最小支持度发生变化时,通过执行更新操作,实现频繁序列树的更新,利用深度优先遍历频繁序列树找到序列数据库中所有的序列模式。实验结果表明,与IncSpan算法和PrefixSpan算法相比,该算法的挖掘效率较高。     

一种有效的序列模式增量式更新方法

针对序列模式增量式更新挖掘算法产生大量候选项集以及多次扫描数据库的问题,提出了一种有效的增量式更新算法ESPIA,该算法利用基于2-序列矩阵挖掘算法ESPE对原数据库和增加数据库一次扫描产生序列模式,通过对频繁模式和非频繁模式进行相应的剪枝减少了序列的比较和扫描次数,降低了算法时间和空间复杂度,实验证明该算法是有效和准确的。     

基于频繁序列树的交互式序列模式挖掘算法

为了减少在序列模式挖掘过程中由于重复运行挖掘算法而产生的时空消耗,提出了一种基于频繁序列树的交互式序列模式挖掘算法(ISPM). ISPM算法采用频繁序列树作为序列存储结构,频繁序列树中存储数据库中满足频繁序列树支持度阈值的所有序列模式及其支持度信息.当支持度发生变化时,通过减少本次挖掘所要构造投影数据库的频繁项的数量来缩减投影数据库的规模,从而减少时空消耗.实验结果表明,ISPM算法在时间性能上优于PrefixSpan算法和Inc-Span算法     

序列模式增量式更新的研究

序列模式挖掘是数据挖掘领域中十分重要的研究课题.目前已有许多算法用于序列模式的挖掘,但在序列模式增量式更新方面的研究还比较少,针对这种情况提出了序列模式增量式更新的挖掘算法SPIU.SPIU算法充分利用了原有的挖掘结果,并对产生的候选频繁序列进行剪枝,有效地减小了候选频繁序列的大小,从而很好地改善了挖掘效率.测试结果表明SPIU算法是正确和高效的,另外算法还具有很好的扩放性.     

最小频繁闭树的增量式更新算法

针对树挖掘算法产生大量频繁子树和树数据库随时间变化的问题,提出最小频繁闭树增量式更新算法以及增量式更新策略,能充分利用已有挖掘知识,无须重新运行树挖掘算法,并且只需进行一次数据库扫描操作。给出一种候选子树剪枝方法,能减少树同构判别次数,有效提高算法的运行效率。通过大量实验结果表明,该算法有效可行且效率较高。     

基于FP_tree的频繁项目集增量式更新算法

对频繁项目集的更新问题进行研究,提出一种基于频繁模式树的频繁项目集增量式更新算法。充分利用已有挖掘结果,有效解决最小支持度和事务数据库同时发生变化时相应频繁项目集的更新问题。在事务数据库变化同时包括增加和减少的情况下,对算法性能进行分析与测试,结果证明该算法高效可行。     

基于频繁模式树的关联规则增量式更新算法

研究了大型事务数据库中关联规则的增量式更新总是，提出了一种基于频繁模式树的关联规则增量式更新算法，以处理最小支持度或事务数据库发生变化后相应关联规则的更新问题，并对其性能进行了分析。     

金融时间序列频繁模式挖掘算法①

针对金融时间序列数据库信息,提出一种时间序列频繁模式自动发现算法,该算法首先构造投影树,然后采用深度优先策略遍历投影树,挖掘出所有最长频繁模式,实验结果表明,该算法成功地挖掘出满足约束的频繁序列,在相同条件、不同支持度情况下,取得了与传统AprioriAll方法相同的规则集,而运行效率优于AprioriAll方法。     

An efficient approach for the maintenance of user behaviours

Mining sequential patterns is to discover sequential purchasing behaviours for most of the customers from a large number of customer transactions. The strategy of mining sequential patterns focuses on discovering frequent sequences. A frequent sequence is an ordered list of the itemsets purchased by a sufficient number of customers. The previous approaches for mining sequential patterns need to repeatedly scan the database so that they take a large amount of computation time to find frequent sequences. The customer transactions will grow rapidly in a short time, and some of the customer transactions may be antiquated. Consequently, the frequent sequences may be changed due to the insertion of new customer transactions or the deletion of old customer transactions from the database. It may require rediscovering all the patterns by scanning the entire updated customer transaction database. In this paper, we propose an incremental updating technique to maintain the discovered sequential patterns when transactions are inserted into or deleted from the database. Our approach partitions the database into some segments and scans the database segment by segment. For each segment scan, our approach prunes those sequences that cannot be frequent sequences any more to accelerate the finding process of the frequent sequences. Therefore, the number of database scans can be significantly reduced by our approach. The experimental results show that our algorithms are more efficient than other algorithms for the maintenance of mining sequential patterns.     

Fast incremental mining of web sequential patterns with PLWAP tree

Point and click at web pages generate continuous data sequences, which flow into the web log data, causing the need to update previously mined web sequential patterns. Algorithms for mining web sequential patterns from scratch include WAP, PLWAP and Apriori-based GSP. Reusing old patterns with only recent additional data sequences in an incremental fashion, when updating patterns, would achieve fast response time with reasonable memory space usage. This paper proposes two algorithms, RePL4UP (Revised PLWAP For UPdate), and PL4UP (PLWAP For UPdate), which use the PLWAP tree structure to incrementally update web sequential patterns efficiently without scanning the whole database even when previous small items become frequent. The RePL4UP concisely stores the position codes of small items in the database sequences in its metadata during tree construction. During mining, RePL4UP scans only the new additional database sequences, revises the old PLWAP tree to restore information on previous small items that have become frequent, while it deletes previous frequent items that have become small using the small item position codes. PL4UP initially builds a bigger PLWAP tree that includes all sequences in the database using a tolerance support, t, that is lower than the regular minimum support, s. The position code features of the PLWAP tree are used to efficiently mine these trees to extract current frequent patterns when the database is updated. These approaches more quickly update old frequent patterns without the need to re-scan the entire updated database.     

基于矩阵的频繁模式挖掘及更新算法

频繁模式挖掘在数据挖掘领域已经有广泛的应用.然而,对于增量更新频繁模式挖掘研究得不是很多.本文提出了一种新颖的增量更新频繁模式树结构(IUNP_Tree),构建它只需要对数据库扫描一次.此外,提出了基于条件矩阵(conditional matrix)的频繁模式挖掘算法(FPBM_Mine)和增量更新算法INUPA,可以有效地处理数据库的增量更新问题.实验表明,该算法是有效的,并且运行效率高于FP-growth算法.     

基于抽样技术的序列模式的维护

维护已发现的序列模式的方法主要有两种：一种是简单地利用已有的挖掘序列模式算法对更新后的整个数据库进行操作,这种方法涉及的数据库中的数据不仅有改变的部分而且有未改变的部分,而未改变的数据数量很大,当更新频率高时,代价是非常大的;另一种方法是根据库中记录数目改变的多少来决定何时对整个数据库进行操作,但是记录数目数据并不能代表序列模式化亦大,因此利用样品抽样的方法来评估序列模式改变的程度,并根据改变的程度决定何时对整个数据库进行操作来更新序列模式,从而较好地解决序列模式维护的问题,能高效地、准确地发现序列模式。     

OSAF-tree--可迭代的移动序列模式挖掘及增量更新方法

移动通信技术和无限定位技术的发展积累了海量的、动态增长的时空数据．利用数据挖掘技术从移动用户的时空行为轨迹当中挖掘用户移动序列模式,在移动通信、交通管理、基于位置服务等领域有着广泛的应用前景．由于移动环境网络资源珍贵、数据量大的特点,传统的序列模式挖掘方法在效率上很难满足需求．OSAF-tree算法基于投影的概念,只需要对数据库进行一遍扫描,就可以很好地处理移动序列模式的挖掘及其增量更新和迭代挖掘问题,这是一个非常高效的算法．与已有的方法相比,OSAF-tree算法在性能和I/O代价等方面都具有明显的优势．     

Discovering Patterns from Large and Dynamic Sequential Data

Most daily and scientific data are sequential in nature. Discoveringimportant patterns from such data can benefit the user and scientist bypredicting coming activities, interpreting recurring phenomena, extractingoutstanding similarities and differences for close attention, compressingdata, and detecting intrusion. We consider the following incrementaldiscovery problem for large and dynamic sequential data. Suppose thatpatterns were previously discovered and materialized. An update is made tothe sequential database. An incremental discovery will take advantage ofdiscovered patterns and compute only the change by accessing the affectedpart of the database and data structures. In addition to patterns, thestatistics and position information of patterns need to be updated to allowfurther analysis and processing on patterns. We present an efficientalgorithm for the incremental discovery problem. The algorithm is applied tosequential data that honors several sequential patterns modeling weatherchanges in Singapore. The algorithm finds what it is supposed to find.Experiments show that for small updates and large databases, the incrementaldiscovery algorithm runs in time independent of the data size.