Master Thesis Defense: Jahnavi Dhananjaya

Speaker: Jahnavi Dhananjaya

Supervisor: Dr. S. Bergler

Examining Committee:
Drs. M. Kersten-Oertel, Y. Yan, V. Haarslev (Chair)

Title:  TripleViz: A Visualization for Triples

Date: Friday, August 12, 2016

Time: 13:00

Place: EV 11.119

ABSTRACT

Most of the data available on the Internet is unstructured. Text mining is the process of automatically extracting information from text. This thesis combines text mining with visualization to develop TripleViz, a lightweight, web-based tool used to process and analyze the text to extract subject-verb-object (SVO) triples, and visualize them as graphs. The SVO triples extracted from documents are visualized using open-source visualization tools Turtled and Gephi. TripleViz extracts noun phrases and visualizes them in either full or thin format to avoid overcrowding on the screen. For the same reason, TripleViz provides an option to select only triples that contain words of interest as provided by the user in form of a word list. Within TripleViz, the user can also view a color-coded output highlighting words from a word list.