reports

When do you tweet the most?
A statistical analysis on UK's Twitter posting time trends
[ HTML Link ] - June 2011
To answer this question, I have used 75 million tweets posted from July to December (2010) and geolocated in the United Kingdom. The general trend derived after performing a simple statistical analysis on this dataset is presented in Figure 1, where you can see the percentage of tweets posted in every hourly interval of a day. Interestingly, one can observe three distinct peaks with increasing frequency, one during lunch break, the next right after work time (for mainstream occupations) and the final one right after dinner time (or right before sleep or during a night out etc.).

MSc Thesis
Weather Talk - Extracting Weather Information by Text Mining
[ Script ] [ Poster ] - September 2008
The main aim of this project was to design and implement a system able to infer the weather state of a location for a specific date by applying Bayesian inference models and statistical analysis on web observations. Additionally, we investigated various linear combinations of probabilistic schemes where traffic information, previous day's weather or a weather prior probability contribute to the final decision. As a final extension, we visualised the weather inference results on a map.

MSc Project Interim Report
[ Review ] [ Plan ] - May 2008
Weather talk takes advantage of the vast amount of information that lies on the World Wide Web. The goal of this project is to use statistical analysis on text documents available online in order to extract information about the weather on a specific location. Documents can include blogs, newsgroups and news articles but not officially weather related web sites or pages. These documents form Weather talk, the input of our project.

Privacy on Web 2.0
[ pdf ] - December 2007 
You already have zero privacy; get over it. These were the exact words of Scott McNealy during the presentation of Sun's Jini. In fact, the insufficient U.S. legislation for web privacy at that time pointed to that direction. Nowadays, the amounts have changed significantly; the world wide web is an extension and in some aspects a replacement of the real world. Internet is a tool for many purposes such as entertainment, shopping, financial management, investing, and socializing. Unfortunately, this massive invasion of Internet in the modern societies exposed weaknesses in their legal systems; their legislation was not ready for this new medium. In this article we focus on the modern aspect of the world wide web, known as Web 2.0, and especially on the legal framework that secures the privacy of its users. The article is organized as follows: firstly, we define Web 2.0 and comment its current general impact; then we focus on user personal data privacy on Web 2.0 and we present possible ways of invoking his privacy; afterwards we draw our attention on web privacy legal frameworks in the United States (hereinafter U.S.) and the European Union (hereinafter EU); in the end, we refer to ways able to enhance user's privacy.

A Report on Timing Attacks
[ pdf ] - December 2007
A side channel attack tries to exploit specific properties of the implementation and operating environment of a cryptosystem rather than its mathematical specification. Timing attacks are a subclass of side channel attacks where the attacker tries to break an encryption algorithm by using information about the execution times of its encryption or decryption queries. In general, a timing attack tries to exploit private information from a system by timing specific system's operations. In this report, a vast amount of different types of timing attacks is described.