The Graphics, Multimedia and GIS Laboratory of the Computer Engineering & Informatics Department of the University of Patras was established in 1994. The educational and research work of the laboratory is being executed within the framework of the Department's Computer Software Section. Read more...


Mr. Viennas presented his PhD Thesis

Mr. Viennas presented his PhD Thesis at 12/07/2017

Title: Searching, Modeling and Evaluation of Applications and Data in the World Wide Web



Mrs. Mparou presented her MSc Thesis

Mrs. Mparou presented her MSc Thesis at 11/07/2017

Title : Digital Advertising Campaigns: Sponsored Search Auctions and Programmatic Advertising

Summary : Online advertising is a kind of promotion through the Web to instantly display messages that are meant to attract customers. The advantage of this is the immediate publication of information and content that is not limited to place or time. It is possible to appear on relevant websites and next to the search listings with relevant keywords. Search engines use auctions where the available advertising space next to the search results is bided for a certain amount of time, and each advertiser has the option of bidding. The highest price offer also ensures that this ad is displayed. In this diploma thesis we will study the development of advertising over the years focusing on two aspects of advertising: Search advertising and Programmatic advertising. By following data mining techniques, we will analyze the Yahoo’s Search dataset (including the bid prices of advertisers who participated in Yahoo Search Marketing auctions for the top 1000 keywords in the 2012-2013 period). Finally, in order to clarify the specifics of the dataset, with algorithms provided by the Weka tool and following classification methods, we will make substantiated conclusions about the algorithms that prevail.

Mr. Pithamitsis presented his Diploma Thesis

Mr. Pithamitsis presented his Diploma Thesis at 10/07/2017

Title:  Analysis of  Twitter’s Streaming Data using Apache Flink

Abstract: Nowadays, data analysis is a key part of the functioning of many aspects of our everyday life. Data analysis provides at analysts and decision-makers the ability to act on based on existing results and data. Using the results of the analysis, they have the ability to act more securely and to be more confident that their actions will have better results. However, the analysis is often either slow or its results are incorrect. The Apache Flink comes to solve this problem.
This diploma thesis aims to present the benefits of the Apache Flink framework over the currently used frameworks such as Hadoop and Spark. Apache Flink is a new hybrid processing framework that can do both batch and stream processing. Its ability to process every type of data as streaming data makes it one of the fastest data analysis tools. In addition, Flink, unlike other stream processing frameworks, provides results that are accurate, even in the case of out-of-order or late-arriving data.

Mrs. Christidi presented her Diploma Thesis

Mrs. Christidi presented her Diploma Thesis at 10/07/2017

Title: Prediction of the indicators of the stock market through sentimental data analysis from Twitter.

Abstract: The subject of this diploma thesis is sentiment analysis in Twitter data, using methods of machine learning. Sentiment analysis is the automated process for extracting information about the sentiment polarity of a given body of text and is often, alternatively referred to, as opinion mining. It is a field of study that currently attracts a lot of academic attention due to the impact of social networks in our everyday life, the automated way it offers for analyzing the written information hugely available in web sources, and also because of the substantial progress being made in the last years, in the fields of machine learning, artificial intelligence and deep learning.

Mr. Papanikolaou presented his Diploma Thesis

Mr. Papanikolaou presented his Diploma Thesis at 10/07/2017

Title: Health Care Analytics Using Hadoop

Summary: Nowadays, every minute 1.7 billion bytes of data are generated, equivalent to 360,000 DVDs, that is, more than 6 megabytes of data per person per day, and this rate rises by 40% per year. So today, especially data analysis in the health sector is a cutting edge in the research as well as the need for the use of these data to generate automatic conclusions. For this purpose, the development of technologies in the management and analysis of large data and the development of technologies in data mining plays an important role.

You are here: Home Home News