For the spell correction task, vocabulary based methods have been replaced with methods that take morphological and grammar rules into account. However, such tools are fairly immature, and, worse, non-existent for many low resource languages. Checking only if a word is well-formed with respect to the morphological rules of a language may produce false negatives due to the ambiguity resulting from the presence of numerous homophonic words. In this work, we propose an approach to detect and correct the "de/da'' clitic errors in Turkish text.
Beginning of every semester, students and instructors try to complete registration schedule of the semester in three days. This situation is not a user-friendly method. People get into difficulty because of the registration system's impracticability.
From the student point of view, the insufficient time of sending the academic program for approval and also waiting for the answer from consent requests become a tremendous problem.
The aim of this project is to create a web based application which provides easy access to drug target interactions, or in a more general sense, protein ligand interactions.
A geostamp is a record of a geographic location with a timestamp. An Ethereum blockchain based data market will be developed where observers will be able to submit geostamps of moving entities and get paid for their usage by others.
In this project, we are interested with the problem of aligning features of human motions and gestures. There are variations in subjects’ characteristics, style, and speed in hand, body, and facial gestures. We try to minimize effects of those variations by a pre-alignment step. Canonical Time Warping (CTW) and Generalized Canonical Time Warping (GTW) are claimed to give better results for human motion alignment than alternative methods. Our aim is to analyze mathematical foundations of GTW, fine tune its parameters for sign language alignment, and find the best setup for GTW to work.
ISPs today offer 100 Mbps to modems, but people never seem to get that through their WiFi. About half of the 10-20% ISP churn is due to performance dissatisfaction, and the leading cause of that is poor WiFi. This costs an ISP like Comcast a billion dollars each year. They desperately need to own home WiFi the way they do the rest of their network, yet all the new technology being built is for “faster” WiFi, not “controlled” WiFi. Today numerous devices connect to internet via wifi. Most of the consumer products do not even have ethernet jack.
The object of my dissertation is identifying the ways to exploit vulnerabilities to defeat the security features of system components used in Boğaziçi University. The goal of penetration testing will be determining whether and how a malicious user can gain unauthorized access to assets that affect the fundamental security of the systems.
Our aim is developing a text mining system consisting of BioC-compatible modules integrated together to assist biocurators.We contributed to the system by developing a module for identifying the passages that describe experimental methods for physical PPIs. Our approach is based on tf-rf and word2vec techniques.
Digital Humanities is an area where computing and humanistic disciplines intersect. In this project, we intersect literature with computing by visualizing Nazım Hikmet’s poetry. We developed a visualization tool to explore the distinctive style of Nazım Hikmet, to examine the changes in his style over the years, and to enable scholars to query aspects of his oeuvre selectively. We combined text-mining methodologies and interactive visualizations for this purpose. A database is created from Nazım’s entire work to store and query both content and structure. We parsed the content,