Mining Sequential Access Pattern with Low Support From Large Pre-Processed Web Logs

Finding Frequent Sequential Pattern (FSP) is an important problem in web usage mining. In this paper, we systematically explore a pattern-growth approach for efficient mining of sequential patterns in large sequence database. The approaches adopts a (divide and conquer) pattern-growth principle as follows: Sequence databases are recursively projected into a set of smaller projected databases based on the current sequential pattern(s), and sequential patterns are grown in each projected databases by exploring only locally frequent fragments. Our proposed method combines tree projection and prefix growth features from pattern-growth category with position coded feature from early-pruning category, all of these features are key characteristics of their respective categories, so we consider our proposed method as a pattern growth / early-pruning hybrid algorithm that considerably reduces execution time. These approaches were implemented in hybrid concrete method using algorithms of sequential pattern mining.

Although efficient identification of user access sessions from very large web logs is an unavoidable data preparation task for the success of higher level web log mining, little attention has been paid to algorithmic study of this problem. In this paper we consider two types of user access sessions, interval sessions and gap sessions. We design two efficient algorithms for finding respectively those two types of sessions with the help of new data structures. We present both theoretical and empirical analysis of the algorithms and prove that both algorithms have optimal time complexity.KeywordsAssociation RuleUser AccessUser NodeAccess RecordInterval SessionThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

We present a data visualization approach that allows users to interactively analyse sequential movement patterns and automatically extract rules for marine animal behaviour. This enhances the ability of oceanic experts to understand marine ecosystems. Traditional time-series analysis, basic statistical methods, and single-view visualizations often fail to capture the intricate spatial and temporal dynamics of marine ecosystems. Our visual analytics approach, that leverages sequential pattern mining and rule mining techniques, integrates SPMF's Vertical Mining of Maximal Sequential Patterns (VMSP) algorithm and T-Rule Growth algorithm to efficiently extract and visualize movement patterns and behavioural rules of marine animals. The system also incorporates automated pre-processing to clean and prepare large datasets, reducing the time and effort required for manual handling. The visualization presents oceanic data from detection stations, allowing researchers to discover patterns in the movements of marine species. Our interface also provides geographical visualization of selected patterns or rules and helps users to navigate and find the relevant information. Key principles of our model include multi-view visualizations that provide comprehensive insights by displaying data from various perspectives simultaneously, and geo-spatial time-varying data visualization that captures and represents the dynamic movements of marine species. This approach allows domain experts to interactively explore and analyse sequential patterns and rules, enhancing their ability to make informed decisions. Additionally, feedback received, and insights gained from specialized professionals was used to determine if this approach would be a beneficial complement to existing practice.

The data mining techniques help discover hidden knowledge from a huge database. In the pattern mining field, the main goal is to discover interesting patterns in large databases. The sequential pattern mining technique is specialized for discovering sequential patterns with only one measure called support. It is not sufficient and misleading for the user. Sequential rule mining is a good solution that takes another measure into an account called confidence. This paper presents a comparative analysis between two sequential rule mining algorithms, namely non-redundant with dynamic bit vector (NRD-DBV), and TRuleGrowth algorithm. The study clarifies the execution time, the number of rules, and the memory usage for each algorithm. In addition, exposure to the most proper field for each algorithm to achieve elevated efficiency.

The process of component security testing can produce massive amounts of monitor logs. Current approaches to detect implicit security exceptions (those which cannot be identified by visual inspection alone) compare correct execution sequences with fixed patterns mined from the execution of sequential patterns in the monitor logs. However, this is not efficient and is not suitable for mining large monitor logs. To enable effective mining of implicit security exceptions from large monitor logs, this paper proposes a method based on improved variable-length sequential pattern mining. The proposed method first mines the variable-length sequential patterns from correct execution sequences and from actual execution sequences, thus reducing the number of patterns. The sequential patterns are then detected using the Sunday string-searching algorithm. We conducted an experimental study based on this method, the results of which show that the proposed method can efficiently detect the implicit security exceptions of components.

Web mining is one of the mining technologies, which applies data mining techniques in large amount of Web data to improve the Web services. Web traversal pattern mining discovers most of the users' access patterns from Web logs. This information can provide the navigation suggestions for Web users such that appropriate actions can be adopted. However, the Web data will grow rapidly in the short time, and some of the Web data may be antiquated. The user behaviors may be changed when the new Web data is inserted into and the old Web data is deleted from Web logs. Therefore, we must re-discover the user behaviors from the updated Web logs. However, it is very time-consuming to re-find the users' access patterns. Hence, many researchers pay attention to the incremental mining in recent years. The essence of incremental mining is that it utilizes the previous mining results and finds new patterns just from the inserted or deleted part of the Web logs such that the mining time can be reduced. In this paper, we propose an efficient incremental Web traversal pattern mining algorithm. The experimental results show that our algorithm is more efficient than the other approaches

Web mining is one of the mining technologies, which applies data mining techniques in large amount of web log data. Web navigational mining discovers users’ access patterns from web logs. This information can be used to identify the behavior of the web user. However, the web data will grow rapidly in the short time, and some of the web data may be antiquated. The user behavior may be changed when the new web data is inserted into and the old web data is deleted from web logs. Therefore, the user behavior must be re-discovered from the updated web logs. However, it is very time-consuming to re-find the users’ access patterns. Hence, many researchers pay attention to the incremental mining, which utilizes the previous mining results and finds new patterns just from the inserted or deleted part of the web logs such that the mining time can be reduced. The present paper proposes an efficient incremental web navigational mining algorithm for discovering web navigational patterns when the user sequences are inserted into and deleted from original database. It avoids re-finding the original web navigational patterns and re-counting the original candidate sequences. It uses lattice structure to keep the previous mining results such that just new candidate sequences need to be computed. Hence, the web navigational patterns can be obtained rapidly when the navigational sequence database is updated. Besides, maximal web navigational patterns can also be obtained easily by traversing the lattice structure. The experimental results show that the present algorithm is more efficient than the other approaches.

A parallel rough set based dependency calculation method for efficient feature selection

The automated quantification of different cell structures available in histopathological images is a challenging task due to the presence of complex background structures. Moreover, the tissues of different categories, namely epithelium tissue, connective tissue, muscular tissue, and nervous tissue have heterogeneous structure which limits the applicability of an algorithm to only a single class of tissue for the quantification analysis of histopathological images. Therefore, this paper introduces a novel method for categorization of histopathological images into the respective tissue category before quantification analysis. The proposed method uses SIFT method for feature extraction which are further processed by gravitational search algorithm to obtain optimal bag-of-visual-words. Moreover, support vector machine is trained on these bag-of-visual-words to classify the images into respective categories. The experimental results show that the proposed method outperforms the traditional K-means-based method for histopathological image classification.

The 'Cold Start' problem, characterized by insufficient transaction history leading to inefficient personalization, represents one of the frequent challenges encountered in e-commerce systems. This issue, along with data sparsity resulting from limited product interactions, further complicates the reliability of conventional recommendation engines. The objective of this research is to design a novel hybridized recommendation system that enhances both security and suggestion accuracy by dynamically adapting to user interactions in digital environments. By leveraging contextual information and sequential user behavior patterns, the proposed method addresses gaps left by traditional recommender systems. A research methodology is employed that integrates context awareness and sequential pattern mining, combining conventional and advanced recommendation techniques within a unified framework. This hybrid strategy tackles challenges posed by Cold Start scenarios while optimizing the use of scarce user interaction data. Preliminary findings indicate that the hybrid approach significantly improves recommendation performance, particularly in settings with limited historical data. By utilizing contextual and sequential information, the proposed system outperforms existing methods in terms of accuracy and personalization. In conclusion, the study demonstrates the effectiveness of hybrid recommendation systems in addressing fundamental e-commerce challenges such as data sparsity and the Cold Start problem. The adaptability of the approach enables accurate predictions without requiring extensive user history. Future research will focus on the scalability of this strategy across diverse e-commerce platforms to deliver more individualized user experiences.

Sequential pattern mining and temporal association rule mining techniques are established techniques to mine insights into the data collected by smart home environments and ambient living systems. Local process models are a recent technique that uses constructs from the field of business process modeling to represent frequent patterns that go beyond sequential patterns and can additionally express rich ordering relations that include concurrent execution, choices, and repetition. To gain insight into the behavior within smart home environments using local process models, it is vital to focus on patterns where the activities that are described by the pattern are truly related. We show on real-life smart home data that existing techniques for mining local process models fail to do so, and address this problem by proposing novel techniques for constraint-based local process model mining, focusing on event gap constraints and time gap constraints. We observe that in order to get an accurate support count of a local process model, it is sufficient to only consider specific parts of the datasets instead of the full dataset, which allows to speed up the counting of the support of a pattern. We provide a novel algorithm to extract those relevant parts of the data for support counting. We evaluate our approach on a collection of real-life smart home data sets. We show that more insightful local process models can be mined when applying such constraints, and show the novel algorithm allows to mine them efficiently.

In this paper we present a frequent sequence pattern mining-based algorithm used for network intrusion detection, which is an application and extension of the SPADE algorithm. It is based on the idea that much behavior on the network appears as sequences of activities, according to the sequence patterns we computed, we can construct the intrusion rule base and legal action rule base, then we can detect known and novel intrusion activities by rule matching. In addition, when the system is running, we use an incremental sequence pattern mining algorithm to complement the rule library in order to avoid re-executing the algorithm on the entire dataset, thereby reducing execution time. The experimental results indicate that this algorithm is efficient enough to meet the needs for active detection of intrusion. Compared with most existing methods used in commercial systems which are built using purely knowledge engineering approaches, our algorithm is more intelligent and adaptive.

Previous studies have shown that emotion is crucial in student learning. However, most studies in the e-learning environment have yet to consider emotion as part of learning that could lead to successful learning. Thus, this study explored the relationship between student emotion state, emotion sequences, and student learning experience. A preliminary data collection was conducted to explore the relationship between emotional experience and student learning experience, which involved 16 students. Students were asked to learn a programming subject in an e-learning environment. E-learning is designed to store the students' emotional experience and activity during learning. The sequential pattern mining technique was used to extract the data, exploratory data analysis was conducted to visualize the emotional trajectory during the learning process, and regression analysis was used to explain the relationship between students' emotional learning experiences. The results showed that emotional experience might affect student experience in learning. In one-sequence emotion, all emotion states contributed to the learning experience with p-values < 0.01 except for neutral and disgust with p-values < 0.05. The one-sequence emotion model shows R-squared = 0.585; Adj. R-squared = 0.734; F-statistic = 6.920; Prob (F-statistic) = 0.00702. Meanwhile, in two-sequence emotion, none of the emotion sequences contributed to the student learning experience. Lastly, three-sequence emotion models also showed that most sequences did not influence student learning experience. The only sequence of emotions that influenced the student learning experience was surprise-neutral-surprise. These results suggest that emotion should be considered in learning design as it can influence student experience.

The application of location-aware technology has increasingly accumulated a large amount of user spatiotemporal data, which gives us the opportunity to analyze users' interests and behavior patterns. This paper proposes a new method for constructing user's mobility profiles and measuring their similarity. First, a new method of using the categories of points of interest (POI) to represent the semantics of a geographical area (stay region) where a user stays for a long period of time is proposed. Second, the sequential pattern mining technique is applied to extract the user's mobility profiles. Then, the user similarity is computed based on the maximum benefit assignment problem. Finally, a comparison experiment with the existing works on the real dataset verifies the effectiveness of the proposed approach.

We study the relationships between affective states and sequences and learner achievement using sequential pattern mining techniques. We found, in accordance with prior research, that boredom is an undesirable state that is both persistent and detrimental to learning. We also found that confusion punctuated with periods of engaged concentration contributes to learning. However, confusion alone has a negative impact on student achievement, possibly indicating that students are stuck. These results shed light on past results finding inconsistent relationships between confusion and learning.