Summary Word embedding approaches represent data sequences to handle their contextual meaning in the NLP tasks. Nowadays, there is an emerging need to understand the user behavior patterns over navigational clickstream data. However, representing the URL data sequences utilizing existing embedding approaches to cluster users' behavior with unsupervised machine learning tasks is a challenging task. This study introduces the Patter2Vec embedding approach using a representation vector to construct contextual, precise, and interpretable clusters over the hidden and popular navigational patterns. To test the usability of the proposed representation in clustering tasks, we conduct an experimental study, which indicates that Pattern2Vec outperforms existing embedding approaches.

The user interface (UI) is the software component with which the end users interact the most in the mobile and web world today. Therefore, when evaluating end-user satisfaction with a software, besides the functional features, the possibilities provided by the UIs are significant. End-user interfaces offer several capabilities today, necessitating a complex structure. Examples include interfaces developed in technologies such as Google Flutter and React. Therefore, we foresee a need for UI testing business workflows allowing comprehensive testing of UIs to improve customer satisfaction. Within the scope of this research, we propose an end-to-end approached automated UI testing business workflow for testing UI components on websites with intense user actions.

User interface (UI) testing is a necessary process to evaluate whether the developed web applications meet the software requirements defined by the end-users. This study proposes an e-business workflow that automates UI testing utilizing Hidden Markov Models. The proposed workflow is designed to generate test scripts based on user browsing behaviors. In particular, it is able to generate browsing patterns that are unseen in the historical web usage data. To demonstrate the usability of the proposed testing framework, we provide a prototype implementation. Furthermore, to facilitate testing of the prototype implementation, we conduct experimental studies that investigate the quality of newly generated test scripts. In order to achieve this, we analyze the similarity between the actual browsing behaviors and the newly generated browsing behaviors. The results show that the proposed testing workflow is able to learn from the user-browsing behaviors extracted from the historical web usage log data. The results also show that the workflow is able to generate new and unseen user-browsing behaviors that have high similarities with the actual user-browsing behaviors. In turn, this indicates that the proposed system is able to generate high-quality new test scripts that can provide comprehensive testing of web applications.

With the advent of web 2.0, web application architectures have been evolved, and their complexity has grown enormously. Due to the complexity, testing of web applications is getting time-consuming and intensive process. In today’s web applications, users can achieve the same goal by performing different actions. To ensure that the entire system is safe and robust, developers try to test all possible user action sequences in the testing phase. Since the space of all the possibilities is enormous, covering all user action sequences can be impossible. To automate the test script generation task and reduce the space of the possible user action sequences, we propose a novel method based on long short-term memory (LSTM) network for generating test scripts from user clickstream data. The experiment results clearly show that generated hidden test sequences are user-like sequences, and the process of generating test scripts with the proposed model is less time-consuming than writing them manually.

User behavior can be identified by clustering the web-site navigational patterns expressed by clickstream sequences. Representation of clickstream sequences is yet a challenging problem. The main difficulty is related to the fact that one needs to represent clickstream sequences, which mainly consists of dynamic URLs, in such a way that traditional clustering algorithms are applicable to group together these sequences. In this study, we present the application of embedding approaches to represent click-stream data sequences, to enable machine learning algorithms learn the users' navigational behaviors on web-sites. By utilizing embedding representation, we propose an algorithm that takes clickstream data as input and creates clustered sequential patterns. We discuss the details of different representation algorithms and present its evaluations. We investigate the accuracy in finding the hidden clustered data sequences within the clickstream data. The results show that Word2Vec, representation method that can lead to high quality clustering of user navigational patterns.

In functional User Interface testing, test scenarios are written with respect to the requirements that are specified by test analysts. Usually, a test analyst focuses on base URLs and HTML components while collecting requirements of User Interface test scenarios. A base URL is essentially a unit segment of large scale graph data. It has mostly dynamic shape and is used to navigate pages amongst application's pages. We argue that even though dynamic URLs have additional important information about the content of the page, they are not being utilized in generating User Interface test scenarios. In this study, we address this lack of capability and focus on the development of a methodology that can support the usage of large-scale dynamic URL datasets in UI test script generation. Our proposed methodology is designed as an add-on tool that can be used on the top of the existing UI test automation tools to improve testing quality. We introduce a higher quality testing methodology to make the results more accurate, and we discuss the proposed methodology and give an overview of the implementation details followed by the evaluation results. We perform various performance evaluations to investigate how well the proposed algorithms scale under increasing data sizes. The results are promising and show the usability of the proposed methodology.

Due to prevalent use of sensors and network monitoring tools, big volumes of data or “big data” today traverse the enterprise data processing pipelines in a streaming fashion. While some companies prefer to deploy their data processing infrastructures and services as private clouds, others completely outsource these services to public clouds. In either case, attempting to store the data first for subsequent analysis creates additional resource costs and unwanted delays in obtaining actionable information. As a result, enterprises increasingly employ data or event stream processing systems and further want to extend them with complex online analytic and mining capabilities. In this paper, we present implementation details for doing both correlation analysis and association rule mining (ARM) over streams. Specifically, we implement Pearson-Product Moment Correlation for analytics and Apriori & FPGrowth algorithms for stream mining inside a popular event stream processing engine called Esper. As a unique contribution, we conduct experiments and present performance results of these new tools with different tumbling and sliding time-windows over two different stream types: one for moving bus trajectories and another for web logs from a music site. We find that while tumbling windows may be more preferable for performance in certain applications, sliding windows can provide additional benefits with rule mining. We hope that our findings can shed light on the design of other cloud analytics systems.