Yehia Taher

The Butterfly: An Intelligent Framework for Violation Prediction within Business Processes

Recent research initiatives in the domain of business process management such as process intelligence, monitoring, and mining have shown significant results in automated process environments. However, such techniques fall short to provide efficient solutions and support for non-fully automated business processes i.e., processes embodying dynamic, continuous, and manual activities such as in logistics. More precisely, things turn to be very challenging when it comes to consider the monitoring and violation predictions through-out context-dependent and manual processes. Unlike current initiatives that mainly focus on the model of the process as a whole, we shift in this work towards instance-based and specific processing for each activity. We showcase a contextualized template-driven framework called the buttery, along side its architecture that could address the needs of continuous monitoring and prediction. Satisfactory results from evaluations on real data demonstrate the effectiveness of our framework.

Complex event processing for the non-expert with autoCEP: demo

The inference mechanisms of CEP engines are completely guided by rules, which are specified manually by domain experts. We argue that this user-based rule specification is a limiting factor, as it requires the experts to have technical knowledge about the CEP language they want to use, it restricts the usage of CEP to merely the detection of straightforward situations, and it restrains its propagation to more advanced fields that require earliness, prediction and proactivity. Therefore, we introduce autoCEP as a data mining-based approach that automatically learns CEP rules from historical traces. autoCEP requires no technical knowledge from domain experts, and it also shows that the generated rules fit for prediction and proactive applications. Satisfactory results from evaluations on real data demonstrate the effectiveness of our framework.

Automatic Learning of Predictive CEP Rules: Bridging the Gap between Data Mining and Complex Event Processing

Due to the undeniable advantage of prediction and proactivity, many research areas and industrial applications are accelerating the pace to keep up with data science and predictive analytics. However and due to three well-known facts, the reactive Complex Event Processing (CEP) technology might lag behind when prediction becomes a requirement. 1st fact: The one and only inference mechanism in this domain is totally guided by CEP rules. 2nd fact: The only way to define a CEP rule is by writing it manually with the help of a human expert. 3rd fact: Experts tend to write reactive CEP rules, because and regardless of the level of expertise, it is nearly impossible to manually write predictive CEP rules. Combining these facts together, the CEP is---and will stay--- a reactive computing technique. Therefore in this article, we present a novel data mining-based approach that automatically learns predictive CEP rules. The approach proposes a new learning algorithm where complex patterns from multivariate time series are learned. Then at run-time, a seamless transformation into the CEP world takes place. The result is a ready-to-use CEP engine with enrolled predictive CEP rules. Many experiments on publicly-available data sets demonstrate the effectiveness of our approach.

Fraud Data Analytics Tools and Techniques in Big Data Era

Fraudulent activities (e.g., suspicious credit card transaction, financial reporting fraud, and money laundering) are critical concerns to various entities including bank, insurance companies, and public service organizations. Typically, these activities lead to detrimental effects on the victims such as a financial loss. Over the years, fraud analysis techniques underwent a rigorous development. However, lately, the advent of Big data led to vigorous advancement of these techniques since Big Data resulted in extensive opportunities to combat financial frauds. Given that the massive amount of data that investigators need to sift through, massive volumes of data integrated from multiple heterogeneous sources (e.g., social media, blogs) to find fraudulent patterns is emerging as a feasible approach.

Automatic learning of predictive rules for complex event processing: doctoral symposium

The inference mechanisms of CEP engines are completely guided by rules, which are specified manually by domain experts. We argue that this user-based rule specification is a limiting factor, as it requires the experts to have technical knowledge about the CEP language they want to use, it restricts the usage of CEP to merely the detection of straightforward situations, and it restrains its propagation to more advanced fields that require prediction and proactivity. Therefore, we introduce autoCEP as a data mining-based approach that automatically learns predictive CEP rules from historical traces. More precisely, we include our novel method that is capable of learning rules and handling events coming from one source, and then we elaborate our vision on how to extend autoCEP to deal with simultaneous events coming from multiple sources.

Towards Rich Sensor Data Representation - Functional Data Analysis Framework for Opportunistic Mobile Monitoring.

The rise of new lightweight and cheap sensors has opened the door wide for new sensing applications. Mobile opportunistic sensing is one type of these applications which has been adopted in multiple citizen science projects including air pollution monitoring. However, the opportunistic nature of sensing along with campaigns being mobile and sensors being subjected to noise and missing values leads to asynchronous and unclean data. Analyzing this type of data requires cumbersome and time-consuming preprocessing. In this paper, we introduce a novel framework to treat such type of data by seeing data as functions rather than vectors. The framework introduces a new data representation model along with a high-level query language and an analysis module. Copyright

ProLoD: An Efficient Framework for Processing Logistics Data

Logistics is a data-intensive industry. The information systems used by logistics companies generate massive volume of data which the companies store to perform different types of analysis. In addition, the advent of Big Data technologies and Internet of Things paradigm have given logistics companies an opportunity to use external data stemming from a wide variety of sources including sensors (e.g., GPS), social media and traffic controlling systems. The logistics companies aim to leverage the power of these external data and perform rigorous analysis in real-time to discover intelligence such as unpredictable delay. However, there are different challenges involved. One of the core challenges is integrating and processing a wide variety of data coming from heterogeneous sources. To the best of our knowledge, there is no off-the shelf solution which can address this challenge. In this paper, we present a framework called ProLoD which performs pre-processing and processing tasks with different types of data. Our framework relies on machine learning algorithms, for processing data; however, we found that the ready to use algorithms are not adequate to guarantee processing efficiency. Therefore, we extended an algorithm called Hierarchical Clustering Algorithm. We evaluated ProLoD by comparing its performance with the HCL algorithm found in the widely-adopted machine learning tool called WEKA. We found that ProLoD is performing reasonably better than WEKA in terms of producing optimal number of clusters.

Fraud Analysis Approaches in the Age of Big Data - A Review of State of the Art

Fraud is a criminal practice for illegitimate gain of wealth or tampering information. Fraudulent activities are of critical concern because of their severe impact on organizations, communities as well as individuals. Over the last few years, various techniques from different areas such as data mining, machine learning, and statistics have been proposed to deal with fraudulent activities. Unfortunately, the conventional approaches display several limitations, which were addressed largely by advanced solutions proposed in the advent of Big Data. In this paper, we present fraud analysis approaches in the context of Big Data. Then, we study the approaches rigorously and identify their limits by exploiting Big Data analytics.

BDLaaS: Big Data Lab as a Service for Experimenting Big Data Solution

Big Data technologies are complex. Building a Big Data ecosystem for deploying and running data products needs specialized skillset. Although, an exhaustive number of technologies have been developed over the last decade, complexity remained an issue for the users. Unfortunately, there is no solution which can reduce heavy manual effort requires configuring and manage complex environment for running data products. There are several platforms which promise an easy to use infrastructure, however, in reality, these platforms are not adequately helpful for many users (data scientists including statisticians and machine learning experts) due to lack of skills in managing system level complexities of Big Data technologies. As a matter of fact, it is difficult for such users to exploit the power of Big Data technologies. Also, learning these technologies is time consuming. For some users, learning complex technologies is nearly impossible. Recently, the notion of container image has drawn an attention because it reduces configuration complexity. A few Big Data technologies have already been containerized. However, the reality is somewhat not the same because containerization requires learning new technologies called Docker. They require manual intervention for configuring images and managing them at runtime. Furthermore, these tasks are cumbersome. In this paper, we present the design and development the initial version of a virtual lab as a service called BDLaaS which will ease building Big Data infrastructure for deploying Big Data solution such as Analytics. We explain how BDLaaS can be used by users without having high-level expertise.

A Context-Aware Analytics for Processing Tweets and Analysing Sentiment in Realtime (Short Paper)

Sentiment analysis has grown to become increasingly important for companies to more accurately understand customer/supplier sentiments about their processes/products and services, and predict customer churn. In particular, existing sentiment analysis aims to better understand their customer’s or supplier’s emotions which are essentially the affirmative, negative, and neutral views of users on tangible or intangible entities e.g., products or services. One of the most prevalent sources to analyse these sentiments is Twitter. Unfortunately, however, existing sentiment analysis techniques suffer from three serious shortcomings: (1) they have problems to effectively deal with streaming data as they can merely exploit (Twitter) hashtags, and (2) neglect the context of Tweets. In this paper, we present SANA: a context-aware solution for dealing with streaming (Twitter) data, analysing this data on the fly taking into account context and more comprehensive semantics of Tweets, and dynamically monitoring and visualising trends in sentiments through dashboarding and query facilities.