Rawane Issa


Boston University
Department of Computer Science
Rafik B. Hariri Institute for Computing and
Computational Science and Engineering

111 Cummington St
Boston, MA, 02215


I am a PhD student in the Department of Computer Science at Boston University and a Fellow at the Hariri Insitute for Computing.

My research interests include the Foundations of Computer Science and (recently) Multi-Party Computation (MPC). I've also recently become interested in Formal Methods as it is a hands on application of Logic.

Im currently working under the direct supervision of Prof. Assaf Kfoury and Prof. Azer Bestavros.

I received my B.E. in Computer and Communication Engineering from the American University of Beirut with a Minor in Mathematics.


Fast MPC Network Distance by Symbolic Optimization of Unrolled Loops.

I am currently extending my work on efficiently and securely performing iterative graph algorithms in the framework of Multi-Party Computation (MPC). This is work is a join effort with Kinan Dak Al-Bab (PhD student at BU).

It includes performing the network distance problem among private companies sub-graphs in a larger graph containing their public connections, by guaranteeing that no information gets leaked at any stage in our computation. We believe that our method is much faster than any other solution found in the literature.

Our solution relies on: only performing the MPC version of the network distance algorithm on a subset of the whole graph; symbolicaly modeling that algorithm in the form of an expression that is later on simplified using the algebraic properties of the operatores found in that algorithm.

  • Scalable Secure Multi-Party Network Vulnerability Analysis via Symbolic Optimization. In Proceedings of WTMC 2017: The IEEE Security and Privacy Workshop on Traffic Measurements for Cybersecurity, San Jose, CA, May 2017.

  • Arabic Temporal Entity Normalization.

    The normalization of extracted temporal entities from a text is an fundamental problem in the natural language processing chain. Projecting expressions onto a timeline, and giving them their correct temporal values allows for a plethora of possibilities in terms of text analysis, ranging from information retrieval to machine translation.
    Thus far, no other work has been done on Arabic temporal entity normalization due to the ambiguities that spring from the morphological richness of the language.

    Under Prof. Fadi Zaraket's supervision, my final year project team and I built a tool that extracts, using ATEEMA (Arabic Temporal Entity Extraction using Morphological Analysis tool), Arabic temporal expressions from texts and normalize them via a time data structure that we have formally specified. Our solution relies on ontologicaly analysing the expressions (via the ontological tree provided by Prof. Mustafa Jarrar), matching them with predefined expert rules written in the form of regular expressions, and finally statisticaly analysing them via Hidden Markov Models.

    We have also collected a rich corpus of newspaper articles that we have tagged and annotated with the corresponding temporal information that they map to, as well as newspaper obituaries. This corpus was used for training and testing our tool and can be later used for other applications.

    Modeling respiratory airways in a circuit environment.

    This project entailed implementing a reduced-order model of the respiratory airways in a circuit environment in order to investigate the transportation of gases from/to the blood in the pulmonary capillary. More specifically, it involved finding a model for the one-dimensional Advection-Diffusion equation with temporally dependent coefficients (This equation describes the propagation and concentration of air in the lungs with respect to time and space).

    Prof. Fadi Karameh, Prof. Issam lakkis (from the mechanical engineering department at AUB) and I, projected the problem onto the neural system because of the similarities we found between our problem and the distribution of membrane potential along passive dendritic trees.

    Our work consisted in correctly modifying and discretizing the neural-model and providing a stability and a boundary conditions analysis of the system. I have also implemented and automated a way of generating a SPICE netlist file which correctly describes the equivalent circuit representing the respiratory airways given certain lung parameters (the branching factor of the lungs for example).

    We are currently finalizing the paper for this project and submiting it for publication.

    Industry Work

  • SAIL at the Hariri Institute for Computing - Software Engineering Fellow (2017).
  • UBILITE - Software Engineering (2015).

  • Publications

  • Scalable Secure Multi-Party Network Vulnerability Analysis via Symbolic Optimization. In Proceedings of WTMC 2017: The IEEE Security and Privacy Workshop on Traffic Measurements for Cybersecurity, San Jose, CA, May 2017.

  • Teaching



    Last Updated: Spring 2017