Iskandar Ishak

Effective mining on large databases for intrusion detection

Data mining is a common automated way of generating normal patterns for intrusion detection systems. In this work a large dataset is customized to be suitable for both sequence mining and association rule learning. These two different mining methods are then tested and compared to find out which one produces more accurate valid patterns for the intrusion detection system.Results show that higher detection rate is achieved when using apriori algorithm on the proposed dataset. The main contribution of this work is the evaluation of the association rule learning that can be used for further studies in the field of database intrusion detection systems.

Mobile Plant Tagging System for Urban Forest Eco-Tourism Using QR Code

The transformation of mobile devices from being a communication devices to be important information resources has changed the way of how activities being implemented including in tourism industry. This is very important for the tourists especially when they visit remote location such as forests. For these tourists, information regarding plants is very important to them and most of the time the tour guides or books will be their source of references. The traditional way of forest tourism lacks the facility for these tourists to obtain information regarding the plants. This paper proposes a system that uses QR codes as tag for tagging the plants that can be scanned by tourists to get further information of the plants. Through this system the visitors only need to use their mobile devices to scan on the plant tag to get its information on their devices. The QR codes used representing the plant ID. Based on these QR codes, the application will query a remote database and retrieved the information of the scanned plant and will be displayed in the visitors mobile devices. Using this system, tourists can get accurate information regarding the plants in the forest through their devices. Also, through this system, they can get information during their visit in the forest quickly without having to depend on tour guides or reference books.

Enhancing Mobile Backend as a service framework to support synchronization of large object

Mobile enterprise applications are primarily developed for existing backend enterprise systems and some usage scenarios require storing of large files on the mobile devices. These files range from large PDFs to media files in various formats (MPEG videos).These files needs to be used offline, sometimes updated and shared among users. In this paper, we enhanced a Mobile Backend as a service (M)BaaS platform that allows resource-limited mobile devices to access large objects from the cloud. Simba [1] is a framework that has support for both table and objects data models, three consistency semantics, resembling strong, causal and eventual consistency. We studied the Simba (MBaaS) framework in detail and found that Simba can be enhanced mainly in three areas of data storage, support for large files and how to handle the scenario of poorly written Mobile apps not to affect the singleton Simba Content Service daemon. Based on our findings, we propose enhanced (M)BaaS framework, a system with improved sync protocol supporting the streaming access to large objects and novel techniques to improve the access speed and sync efficiency for large files for mobile cloud storage services. The Simba client data storage is improved with a storage system capable of supporting both file system and database workloads efficiently using a single, versatile data structure: Variable- Transmission tree (VT-tree), instead of the current implementation of using LevelDB, which is a key value store (KVS) based on a log-structured merge (LSM) tree. The required modifications for the Sync protocol at the client and server side are also discussed.

Deception detection approach for data veracity in online digital news: Headlines vs contents

Veracity is a way to find the truthfulness, availability, accountability and authenticity while deception refers to the way of identifying whether verbal expressions or the overall content is truthful or not. Among the issue in data veracity is the use of deception element in digital news content. Many research have been conducted to address the issue of deception especially in news content they proposed machine learning-based approaches to detect deception in news content. In this paper we compare available deception detection model to improve deception detection accuracy for online digital news veracity. We also proposed a framework to improve deception detection accuracy over digital news portal focusing on headlines. Furthermore, this paper also discussed potential directions for future research in deception of online news.

Classification of Histopathology Images of Breast into Benign and Malignant using a Single-layer Convolutional Neural Network

Breast cancer is known as the second most prevalent cancer among women worldwide, and an accurate and fast diagnosis of it requires a pathologist to go through a time-consuming process examining different captured images under varying magnifications. Computer vision and machine learning techniques are used by many scholars to automate this process and provide a faster and more accurate diagnosis of such cancers, but most of them have utilized handengineered feature descriptors to classify the type of images (whether benign or malignant) based upon. Deep learning techniques have made a significant progress in the world of pattern recognition, image classification, object detection, etc. Convolutional Neural Networks (CNN) -- a special kind of deep learning methods -- are best known for identifying patterns in the images; they try to represent an abstract form of images containing the most salient information needed for distinguishing them from different similar-looking images. The main aim of this paper is to employ CNN for the task of breast cancer classification given an unknown image of the patient for an accurate diagnosis. A new network design is proposed to extract the most informative features from a collection of histopathology images provided by BreakHis database of microscopic breast tumor images. The experimental results carried on 1,995 histopathological images (with a 40× magnifying factor), demonstrated an improved accuracy compared to some prior works, and a comparable performance regarding one of the previous works.

Towards an Enhanced Aspect-based Contradiction Detection Approach for Online Review Content

User generated content as such online reviews plays an important role in customer’s purchase decisions. Many works have focused on identifying satisfaction of the reviewer in social media through the study of sentiment analysis (SA) and opinion mining. The large amount of potential application and the increasing number of opinions expresses on the web results in researchers interest on sentiment analysis and opinion mining. However, due to the reviewer’s idiosyncrasy, reviewer may have different preferences and point of view for a particular subject which in this case hotel reviews. There is still limited research that focuses on this contradiction detection in the perspective of tourism online review especially in numerical contradiction. Therefore, the aim of this paper to investigate the type of contradiction in online review which mainly focusing on hotel online review, to provide useful material on process or methods for identifying contradiction which mainly on the review itself and to determine opportunities for relevant future research for online review contradiction detection. We also proposed a model to detect numerical contradiction in user generated content for tourism industry.

Towards an Enhancement of Organizational Information Security through Threat Factor Profiling (TFP) Model

Information security has been identified by organizations as part of internal operations that need to be well implemented and protected. This is because each day the organizations face a high probability of increase of threats to their networks and services that will lead to information security issues. Thus, effective information security management is required in order to protect their information assets. Threat profiling is a method that can be used by an organization to address the security challenges. Threat profiling allows analysts to understand and organize intelligent information related to threat groups. This paper presents a comparative analysis that was conducted to study the existing threat profiling models. It was found that existing threat models were constructed based on specific objectives, thus each model is limited to only certain components or factors such as assets, threat sources, countermeasures, threat agents, threat outcomes and threat actors. It is suggested that threat profiling can be improved by the combination of components found in each existing threat profiling model/framework. The proposed model can be used by an organization in executing a proactive approach to incident management.

Replacing missing values using trustworthy data values from web data sources

In practice, collected data usually are incomplete and contains missing value. Existing approaches in managing missing values overlook the importance of trustworthy data values in replacing missing values. In view that trusted completed data is very important in data analysis, we proposed a framework of missing value replacement using trustworthy data values from web data sources. The proposed framework adopted ontology to map data values from web data sources to the incomplete dataset. As data from web is conflicting with each other, we proposed a trust score measurement based on data accuracy and data reliability. Trust score is then used to select trustworthy data values from web data sources for missing values replacement. We successfully implemented the proposed framework using financial dataset and presented the findings in this paper. From our experiment, we manage to show that replacing missing values with trustworthy data values is important especially in a case of conflicting data to solve missing values problem.