Bilal Hawashin 

Bilal Hawashin, Ahmad Abusukhon, Ayman Mansour
International Conference on Machine Learning and Data Analysis 2015
Abstract—
This paper proposes an efficient method to
extract user interests for recommender systems. Although
item-item content similarity has been widely used in the
literature, it could not detect certain user interests. Our
solution improves the current work in two aspects. First, it
improves the current recommender systems by detecting
actual user interests. Second, it considers many types of user
interests such as single-term interest, time interval interest,
multi-interests, and dislikes. This extractor would improve
recommender systems in many aspects. Our experiments show
that our proposed method is efficient in terms of accuracy and
execution time.

Ayman M. Mansour1*, Mohammad A. Obaidat2 and Bilal Hawashin3
International Journal of Advanced Computer Research (ISSN (Print): 2249-7277 ISSN (Online): 2277-7970)
Volume-4 Number-4 Issue-17 December-2014
Abstract
Monitoring the health condition of elderly people is a complex problem that involves different medical units and requires continuous monitoring. Besides there is the case if we realistically assume that there does not exist a set of rules that are readily acceptable to all human experts. The parameters used in identifying the medical conditions of a patient are really a vague, subjective measure rather than an objective measure. A more effective system is needed as the electronic patient records become more and more easily accessible in various health organizations such as hospitals, medical centers and insurance companies. These data provide a new source of information that has great potentials in monitoring the health condition of Elderly people. In this paper we have developed a fuzzy inference engine for finding risk factor of elderly People. The reasoning is based on a fuzzy inference system implemented using MATLAB. Fuzzy logic is used to represent, interpret, and compute vague and/or subjective information which is very common in medicine. The Detector is a fuzzy rule-based system. Using clinical information of more than 500 patients treated at the Tafila Technical University Medical Center, we have generated preliminary simulated detection results.

Bilal Hawashin, Farshad Fotouhi, Traian Marius Truta, William I. Grosky

Transactions on Data Privacy 5(1): 297-331 (2012)

Abstract

During the similarity join process, one or more sources may not allow sharing its data with other sources. In this case, a privacy preserving similarity join is required. We showed in our previous work [4] that using long attributes, such as paper abstracts, movie summaries, product descriptions, and user feedbacks, could improve the similarity join accuracy using supervised learning. However, the existing secure protocols for similarity join methods can not be used to join sources using these long attributes. Moreover, the majority of the existing privacy‐preserving protocols do not consider the semantic similarities during the similarity join process. In this paper, we introduce a secure efficient protocol to semantically join sources when the join attributes are long attributes. We provide two secure protocols for both scenarios when a training set exists and when there is no available training set. Furthermore, we introduced the multi‐label supervised secure protocol and the expandable supervised secure protocol. Results show that our protocols can efficiently join sources using the long attributes by considering the semantic relationships among the long string values. Therefore, it improves the overall secure similarity join performance.

Bilal Hawashin, Farshad Fotouhi, Traian Marius Truta

PAIS 2011: Article 6

Abstract

During the similarity join process, one or more sources may not allow sharing the whole data with other sources. In this case, privacy preserved similarity join is required. We showed in our previous work [4] that using long attributes, such as paper abstracts, movie summaries, product descriptions, and user feedbacks, could improve the similarity join accuracy under supervised learning. However, the existing secure protocols for similarity join methods can not be used to join tables using these long attributes. Moreover, the majority of the existing privacy-preserving protocols did not consider the semantic similarities during the similarity join process. In this paper, we introduce a secure efficient protocol to semantically join tables when the join attributes are long attributes. Furthermore, instead of using machine learning methods, which are not always applicable, we use similarity thresholds to decide matched pairs. Results show that our protocol can efficiently join tables using the long attributes by considering the semantic relationships among the long string values. Therefore, it improves the overall secure similarity join performance.

Bilal Hawashin, Farshad Fotouhi, William I. Grosky

IEEE ICDMW 2010: 9-16

Abstract

This paper adopts the use of the diffusion maps method for joining long string values, such as paper abstracts, movie summaries, product descriptions, and user feedback, to improve the performance of the existing similarity join methods. In this work, we showed that using attributes of long string values to detect similar records would significantly improve the overall similarity join performance. Most databases include attributes of long string values, and the existing similarity join methods are not efficient in finding the similarity among the values of these long attributes. In this paper, multiple methods were compared according to their ability in joining long string values semantically.

Ismail Hmeidi, Bilal Hawashin, Eyas El-Qawasmeh

Advanced Engineering Informatics – Elsevier 22(1): 106-111 (2008).

Abstract

This paper reports a comparative study of two machine learning methods on Arabic text categorization. Based on a collection of news articles as a training set, and another set of news articles as a testing set, we evaluated K nearest neighbor (KNN) algorithm, and support vector machines (SVM) algorithm. We used the full word features and considered the tf.idf as the weighting method for feature selection, and CHI statistics as a ranking metric. Experiments showed that both methods were of superior performance on the test corpus while SVM showed a better micro average F1 and prediction time.