Logistic Regression for Detecting Untrustworthy Recommendations in Pervasive Environments
Mohammad Said El-Bashir
Prince Hussein Bin Abdullah College for Information Technology, AL al-Bayt University, Mafraq, Jordan.
Abstract— In recent research, the assessment of the trustworthiness of a certain recommender in pervasive environments is examined upon their previous interactions. However, when it is the initial interaction for a certain user, then recommendations cannot be assessed against trustworthiness. One of the approaches is to refer to a previous interaction of one of the users. However, this approach may give good results but it may also lead into wrong recommendations. In this paper, a method for detecting untrustworthiness in pervasive environments is proposed. After digitizing the data attributes, logistic regression is applied. The data attributes used are the recommendations not users information. The proposed method achieved promising results, which are comparable with other research.
Keywords-Trust model; pervasive environments; logistic regression.
The Challenges of Big Data Visual Analytics and Recent Platforms
Hoda A. Abdelhafez
1. Faculty of Computers & Informatics, Suez Canal University, Ismailia – Egypt
2. College of Computer & Information Sciences, Princess Nourah University, Riyadh, Saudi Arabia
Abeer A. Amer
Sadat Academy for management and Sciences, Alexandria, Egypt
Abstract— Visual Analytics plays an important role in discovering hidden information from massive, heterogenous and streaming data. Visual analytics using visual representations and interactive techniques that are combined with statistical and machine learning methods for analysis process. Big data visual analytics faces many challenges related to technological issues and human cognition. This paper’s aim is to focus on the challenges of big data visual analytics and how the recent platforms including Knime, SAS visual analytics, Arcadia enterprise and TensorFlow could deal with these challenges. It also provides comparison between these platforms. The results show that these platforms can overcome most of the general challenges. TensorFlow is a promising platform for handling challenges related to interaction and user interfaces. Other specific challenges like uncertainty and human cognitive bottlenecks still need more efforts.
Keywords- Big data; Visual Analytics; Challenges; Visual Analytics platforms.
A Content-Based Schema Matching Tool
Dept. of Computer Science and Artificial Intelligence, Faculty of Computer Science and Engineering, Jeddah, Saudi Arabia.
Faculty of Business and Information Technology, University of Ontario Institute of Technology, Oshawa, ON, Canada.
Abstract— Schema matching (SM) is a fundamental task of data integration and data warehousing. Often SM is performed manually which is time consuming and error prone. Furthermore, existing SM tools do not scale well to large schemas. To alleviate these challenges, a novel tool is proposed for automated schema mapping based on the content by matching data entities exclusively based on the content. The resulting topology is convenient to visually explore the relationship among database entities even in large volume. Also, a post-processing algorithm based on data types is proposed for further enhancement clustering results. We present a case study to demonstrate the efficiency and the practicality of the proposed tool.
Keywords- Schema Matching; Schema Mapping; Data Integration; Content-Based Schema Matching; Instance-Based Schema Matching; Visualization.
Citation: Farid Bourennani, Mike Bourque, “A Content-Based Schema Matching Tool”, The World of Computer Science and Information Technology Journal (WSCIT). 2019 Volume 9, Issue 5, pp.22.27.
Mobile Based Attendance System Using QR Code
Hoda Abdelhafez, Maram Alamri, Riyof Alomari, Bayader Alzoman, Rfeef BinSheeha, Ayah Albawardi, Rehab Alzahrani
Information Technology Department, College of Computer & Information Sciences,Princess Nourah University, Riyadh, Saudi Arabia.
Abstract—The mobile attendance systems are used to reduce the time and effort wasted in taking attendance in the colleges. Our research focuses on the creating mobile attendance system using Android studio and database to avoid lots of paper work and prevent data loss. This system applies QR code for each subject, so each faculty member shows the QR code to the students to scan it and make them self-present. The other option is that the faculty member takes attendance by calling the name of the students in class. If the student attends, the faculty member will pull the button ant it will change to green (the student is presence), if the student absent the button remain the same without pulling it. Faculty members can manage and check absences of students. Student can only make attendance for each subject and check their absence hours. The output will be the number of students’ absences weekly or monthly. The developed system helps faculty members to mange attendance easily and efficiently.
Keywords-Mobile App; Android; Attendance; QR code; Faculty.
Citation: Hoda Abdelhafez, Maram Alamri, Riyof Alomari, Bayader Alzoman, Rfeef BinSheeha, Ayah Albawardi, Rehab Alzahrani, “Mobile Based Attendance System Using QR Code”, The World of Computer Science and Information Technology Journal (WSCIT). 2019 Volume 9, Issue 4, pp.17.21.
Fatima T. Al-Khawaldeh
Department of Computer Science, University of York, York, United Kingdom.
Abstract—There are many reasons behind research on speculation and negation: there is a lot of irrelevant (nonfactual) information, and a huge changing with new discovering information may strengthen or weaken previous knowledge. Speculation and negation values are considered as one of the main factors which play an essential role to predict the factuality of event or sentence. Negation reverses the truth of a statement to give the opposition and speculation increase or decreases the uncertainty of statement. Recently, Deep Neural Networks (DNN) have proven better performance to distinguish factual from nonfactual information. Most previous approaches have been dedicated to the English language. To our knowledge, there is no previous developed research to identify the negative or speculative expression for biomedical texts in the Arabic language. This research will develop DNN-based Speculation and negation detection models that able to check claims (negated or speculated sentences) by considering syntactic paths between speculation or negation cues and the remaining words (candidates) in biomedical texts, using Stanford dependency parser. In this paper, the implemented models are evaluated based on the BIOARABIC corpus. Experiments on BIOARABIC corpus show that DNN models achieve a competitive performance and the Attention based Bidirectional Long Short-Term Memory model achieves the best F-scores of 73.55.
Keywords- Arabic NLP; negation; speculation; biomedical (medical and biological); factuality.
Citation: Fatima T. Al-Khawaldeh, “Speculation and Negation Detection for Arabic Biomedical Texts”, The World of Computer Science and Information Technology Journal (WSCIT). 2019 Volume 9, Issue 3, pp.12.16.
Najah Al-shanableh, Mofleh Al Diabat
Department of Computer Science, Al Albayt University, Al Mafraq- Jordan.
Abstract— This research aims to use data mining to predict health care outcomes. We will investigate patterns of multiple chronic conditions (MCCs), or multimorbidity, among the US elderly population. The multimorbidity prediction model, as a general aspect, was not found in the literature, although some researchers have been exploring the risk of developing further chronic conditions after reporting an index disease. Data mining can provide richer results compared to those produced using a statistical approach and greater depth and breadth. It can also help professionals to identify the best time to intervene. In this research, the primary focus was on building disease knowledge using data mining algorithms for MCCs in the elderly. We identified potential morbidity groups using clustering and tested several prediction models on HCUP real data with high accuracy, where the highest accuracy of 99.05% was achieved by Logistic Regression.
Keywords- Multimorbidity; Data mining; Classification; Clustering; Prediction; Chronic Diseases.
Citation: Najah Al-shanableh, Mofleh Al Diabat, “Multimorbidity Prediction Using Data Mining Model”, The World of Computer Science and Information Technology Journal (WSCIT). 2019 Volume 9, Issue 2, pp.7.11.
Customer Segmentation Based on GRFM: Case Study
Sahar Ghoreishi, Industrial Engineering Department K.N.Toosi University of Technology Tehran, Iran.
Keyvan Khandestani, Information Technology Department, Electronic Branch, Islamic Azad University, Tehran, Iran.
Abstract—in the last decades’ firms which have directly or indirectly contact with a customer migrate from product-oriented to be a customer-oriented, hence, some products and customers are not profitable in the same way and some of them bring detriment to the firm. In this regard, firms should recognize loyal, profitable and potential customers with a glance of impressive product which brings added value for them. In order to distinguish profitable customers, they supposed to cluster customers and study their behavior’s group for the sake of having the best investment in the best segment. In this paper, we utilize customize GRFM (Group RFM) to cluster customers based on proposed APC (account-pattern constraint clustering) algorithm. Hence, we calculate the cluster RFM value which could aid the bank to explore both profitable accounts and customers.
Keywords-Component; Data Mining; Constraint Clustering Algorithms; Segmentation; RFM.
Citation: Sahar Ghoreishi, Keyvan Khandestani, “Customer Segmentation Based on GRFM: Case Study”, The World of Computer Science and Information Technology Journal (WSCIT). 2019 Volume 9, Issue 1, pp.1.6.
Mada’ Abdel Jawad, Saeed Salah, Raid Zaghal
Department of Computer Science, Al-Quds University, Jerusalem, Palestine.
Abstract—A Mobile Ad-hoc Network (MANET) is a dynamic single or multi-hop wireless network where nodes are connected wirelessly, and the network is self-configured. Due to the high mobility of nodes, network topology changes more frequently and thus, routing becomes a challenging task. Several routing protocols have been proposed by the researchers for MANETs like the well-known Destination Sequenced Distance Vector (DSDV) and its variants. It is a table-driven routing protocol that was mainly proposed to solve routing loop problems and it performs very well in sparse and low mobility environments. However, it suffers from several performance issues when implemented on high and dense MANETs. A number of modifications of DSDV have been proposed to make it more adaptive and suitable for different environments. In this paper, the performance of DSDV, E-DSDV, I-DSDV, and O-DSDV routing protocols is compared. The performance metrics that were considered in this analysis are packet delivery ratio, throughput, End-to-End delay, and routing overhead. Several simulation scenarios were carried out using the Network Simulator tool (NS3) by varying the number of nodes, pause time and velocity. The simulation results have shown that I-DSDV outperforms the others in low mobility scenarios, whereas O-DSDV has the best performance in high velocity environments.
Keywords-MANET; DSDV; I-DSDV; E-DSDV; O-DSDV Simulation; Network Performance; NS3.
Citation: Mada’ Abdel Jawad, Saeed Salah, Raid Zaghal, “Performance Comparative Study of DSDV, E-DSDV, I-DSDV and O-DSDV MANET Routing Protocols”, The World of Computer Science and Information Technology Journal (WSCIT). 2018 Volume 8, Issue 4, pp.24.31.