30.01.2014 Aufrufe

Abstract-Band - Fakultät für Informatik, TU Wien - Technische ...

Abstract-Band - Fakultät für Informatik, TU Wien - Technische ...

Abstract-Band - Fakultät für Informatik, TU Wien - Technische ...

MEHR ANZEIGEN
WENIGER ANZEIGEN

Erfolgreiche ePaper selbst erstellen

Machen Sie aus Ihren PDF Publikationen ein blätterbares Flipbook mit unserer einzigartigen Google optimierten e-Paper Software.

with a CIG can facilitate implementation and assist in the design of CIGs, but<br />

enabling medical experts to derive insights from this wealth of data is a<br />

challenging task. In this thesis the procedural checking of compliance with an<br />

example guideline (artificial ventilation of newborn infants) is combined with<br />

interactive visualization. The approach suggested in this thesis covers actionbased<br />

checking of compliance with a CIG, three types of actions considering<br />

compliance are defined: invalid actions, valid actions and missing actions.<br />

Visual mappings and encodings for compliance information found in the<br />

course of this work have been integrated into views for patient, treatment and<br />

guideline data. Highlighting and visual abstraction inside views for timeoriented<br />

patient parameters are supported. The plan-execution view, showing<br />

the executed actions during a treatment session, is enriched with encodings<br />

for invalid actions and interval-based representations of repeatedly missing<br />

actions. Furthermore aggregated compliance information is represented visually<br />

inside a view showing the guideline structure. The approach presented in<br />

this thesis is based on the extension of the CareCruiser prototype and implemented<br />

in Java. A low scale evaluation was conducted with a physician to<br />

gather qualitative data and feedback. Results indicate that the visual encodings<br />

for compliance information seem to be intuitive and useful in general,<br />

suggestions and found issues give input for future adaptions and refinements.<br />

Christian Bors<br />

Storage and Visualization of Heterogeneous Data from Online Social Networks<br />

Studium: Masterstudium Medizinische <strong>Informatik</strong><br />

BetreuerIn: Privatdoz. Dr. Edgar Weippl<br />

60<br />

Analyzing the information provided by the users in Online Social Networks<br />

(OSNs) is on the one hand a tedious task but on the other hand people lose<br />

control of the information they have shared with the network, this leaves the<br />

door open for criminals and their counterpart: Forensic investigators. The<br />

context of this thesis is to develop a software tool which facilitates analysis of<br />

Social Snapshots split up into two major modules, the first one being a storage<br />

solution and the other one dealing with visualization and analysis assignments.<br />

Utilizing a database solution we make the usually complex data set easily<br />

accessible and available to be further processed. In terms of visualization we<br />

are focusing on automated graph representations of social networks providing<br />

high usability. This is paired with an export functionality generating comprehensive<br />

reports of a single social snapshot. To store and access such a heterogeneous<br />

set of data efficiently we assess different database concepts and<br />

implement a traditional relational database as well as a modern NoSQL database<br />

to determine an optimal solution regarding performance and accessibility.<br />

An abstract access model delivers queried data to the - visualization as well as<br />

statistics related - processing modules, coverting the underlying physical database<br />

connection. These processing modules are integrated into Gephi, a<br />

comprehensive graph visualization tool, harboring all snapshot analysis related<br />

functionality. Efficient graph representations and clustering methods provide<br />

an easily comprehendible social network overview. After having analyzed a

Hurra! Ihre Datei wurde hochgeladen und ist bereit für die Veröffentlichung.

Erfolgreich gespeichert!

Leider ist etwas schief gelaufen!