Summary

Hany SalahEldeen is an Egyptian born on the 10th of May 1986. In 2008 he graduated with honors from Alexandria University with a Bachelors in Computer Systems Engineering. His graduation project team won the first prize in the graduation projects competition held at the university that year with their project "VOID" which was a Virtual Web-based Integrated Development Environment. Their project was elected to compete nationally in the enterprunerial competition held by the ministry of communication in Egypt. During college, he worked as a Software Development Engineer and trainer in a local company called eSpace Solutions.

After graduation he travelled to Spain and in August of 2009 he received his Masters degree from Universitat Autonoma de Barcelona in Computer Vision and Artificial Intelligence with a thesis entitled: "Colour Naming Using Context-Based Learning through a Perceptual Model". After that he interened at Microsoft Research center in Cairo working on social-based recommendation systems.

In January of 2010 he started his PhD at Old Dominion University in Virginia, USA majoring in Computer Science. In march of the same year he joined Dr. Michael Nelson's research group WS-DL "Web Science and Digital Libraries". During the summer of the same year he travelled to Zurich, Switzerland to intern at Google GmbH working on a project in cooperation between Google MENA team and Google translate team. The following summer of 2011 he joined Microsoft Silicon Valley as an intern as well in the PowerPoint team based in Mountain View California. Currently he is working on his thesis focussing on URL shortening and users intention trying to create a predictive model and apply it in social networks.

Masters Work

Colour Naming Using Context-Based Learning through a Perceptual Model:
Creating a perceptual color naming model through a context-based learning process, utilizing a variant of 3D Sigmoid functions in representing the CIE Lab color space to compute an imitation to human- based perception to homogenous color regions in images and videos. This new model was trained on a large dataset obtained from Google Images and tested against another dataset from the auction website eBay to measure the performance, the model succeeded in exceeding the state of the art approaches significantly.
For further information: Report, Presentation

Doctoral Work

Detecting, Modelling & Predicting User Intention in URL Shortening across Social Networks:

The amount of user generated content on the web has been increasing rapidly in the last couple of years with the increasing popularity of the social networking, online shopping and microblogging websites. Studies have been published focussing on analyzing and modelling the spread and growth of nowadays popular social networking and microbloging services. Inserting the human "user" factor in the equation made this analysis both harder but more realistic. In recent studies user interaction with search engines have been analyzed closely and models immerged describing and even predicting the user's next search, next product recommendation, or even displaying more expected results for the user. Other studies also tried to model and predict user's frustration with a certain search procedure or within a social network for who has the highest influence and where rumors immerged.
URL shortening services appeared with the need to squeeze and beautify long URLs to ease dissemination of links in the social media, especially microblogging services as twitter, and to avoid link breaks in emails. Our focus in this doctoral research is to detect user intentions but in creating, publishing and derefrencing shortened URLs as an alternative web to the web of long regular URIs. The next step is to carefully analyze and examine those intentions to create a parametric/probabilistic model where we will be able to predict user intention upon interacting with those short URLs. Upon this prediction the model can navigate through time to which archived or current version of the resource to provide a better time-based navigation. This will pave the way for many services to be built upon our model to make more use of our memento framework and to create an autonomous system which is intelligent enough to limit the spread of undesired resolved versions.

Publications

Colour Naming Using Context-Based Learning through a Perceptual Model: Hany M. SalahEldeen, R. Benavente, M. Vanrell. Masters Thesis. Universitat Autonoma de Barcelona, 2009.

A Computational Colour Naming Model Trained on Real-Life Images. Hany M. SalahEldeen, R. Benavente, M. Vanrell, New Trends and Challenges in Computer Vision: Progress of Research and Development, page 46--51 - Oct 2009, Universitat Autonoma de Barcelona CVCRD 2009.

How much of the Web is Archived? Scott Ainsworth, Ahmed AlSum, Hany SalahEldeen,Michele C. Weigle, Michael L. Nelson. In JCDL 2011: Proceedings of the 11th ACM/IEEE-CS joint conference on Digital libraries, 2011.

PhD Progress

Coursework
Advanced level course requirement
Required to complete 27 credit hours of post master's coursework. Among which 4 800 level courses taught by at least 3 different professors maintaining at least B grade. At least 18 of the 27 credit hours must be in Computer Science.
Courses Completed

• 3 Credits: 779 Design of Network Protocols. Dr. Wahab -- Spring 2010 Grade: A
• 3 Credits: 851 Introduction to Digital Libraries. Dr. Nelson -- Spring 2010 Grade: A
• 3 Credits: 550 Database Concepts. Dr. Levinestein -- Spring 2010 Grade: A
• 3 Credits: 895 Time on the Web. Dr. Nelson -- Fall 2010 Grade: A
• 3 Credits: 895 Machine Learning. Dr. Zeil -- Fall 2010 Grade: B+
• 3 Credits: 891 Topics of Computer Science. Dr. Nelson -- Fall 2010 Grade: Pass
• 3 Credits: 896 Visual Analytics. Dr. Weigle -- Spring 2011 Grade: A
• 6 Credits: 891 Topics of Computer Science. Dr. Nelson -- Spring 2011 Grade: Pass
• 3 Credits: 891 Topics of Computer Science. Dr. Nelson -- Summer 2011 Grade: Pass
• 3 Credits: 896 Topics of Computer Science. Dr. Nelson -- Fall 2011 Grade: Pass
• 3 Credits: 896 Topics of Computer Science. Dr. Nelson -- Spring 2012 Grade: Pass
• 3 Credits: 595 Web Server Development. Dr. Nelson -- Spring 2012 Grade: A

Percentage Completed: 144.4%
Credit hours Completed: 39
Finished the Core 4 Courses: Done

Dissertation Coursework:
Required to complete at least 24 credit hours of dissertation work.

• 6 Credits: 899 Doctoral Dissertation. Dr. Nelson -- Fall 2011 Grade: Pass
• 3 Credits: 899 Doctoral Dissertation. Dr. Nelson -- Spring 2012 Grade: Pass
• 3 Credits: 899 Doctoral Dissertation. Dr. Nelson -- Summer 2012 Grade: Pass

Percentage Completed: 50%
Credit hours Completed: 12

Examination
Breadth Oral Examination: Passed
(Committee: Dr. Nelson, Dr. Zeil, Dr. Chrisochoides & Dr. Li)
Research Ability Oral Examination: Completed by Exemption (Still applying).
Candidacy Examination: Working on proposal (estimated September).

Remaining:

• 6 Credit hours of dissertation coursework.
• Candidacy exam.
• Dissertation defense.