Bilder vom Neuenheimer Feld, Heidelberg und der Universität Heidelberg

Lehrveranstaltungen
heiCO
Ressourcen	Fachschaft
Studien-FAQ	Technik-FAQ

Knowledge discovery

Kursbeschreibung

Studiengang	Modulkürzel	Leistungs- bewertung
BA-2010	AS-CL	8 LP
Master	SS-CL, SS-TAC	8 LP

Dozenten/-innen	Vivi Nastase
Veranstaltungsart	Hauptseminar
Erster Termin	24.10.2016
Zeit und Ort	Mo, 11:15–12:45, INF 325 / SR 24 (SR)
Commitment-Frist	tbd.

Leistungsnachweis

Paper presentation + course project

Inhalt

The high amount of scientific literature, and data in general, is staggering, and it is unavoidable that in our work we can miss either relevant previous research, or we lack enough knowledge of associated domains from which to draw inspiration and useful methods.

In this course we will look at knowledge discovery from various points of view, depending on what we consider "missed knowledge" to be -- simple "one-step" relations, connections between scientific fields, discovering connections between processes in a sequential manner, and possibly much more.

Kursübersicht

Seminarplan

Datum	Sitzung	Materialien
24.10.2016	missed ...
31.10.2016	lecture	Introduction and Literature Based Discovery
7.11.2016	lecture	Knowledge Graphs and Link Prediction
14.10.2016	lecture	Knowledge Discovery in Wikipedia and Databases List of proposed papers
21.11.2016	paper presentations	(Julian) Diachronic word embeddings reveal statistical laws of semantic change Hamilton et al., ACL 2016 [slides]
28.11.2016	paper presentations	(Jayson) Modeling concept dependencies in a scientific corpus Gordon et al., ACL 2016 [slides] a small annotation exercise
5.12.2016	paper presentations	(Naghmeh) Knowledge base inferencing using bridging entities Kotnis et al., EMNLP 2015 [slides]
12.12.2016	paper presentations	(Shideh) Link prediction based on local random walk Liu & Lu, EPL, 2010 [slides] (Julius) Compositional learning of embeddings for relation paths in knowledge bases and text Toutanova et al., ACL 2016 [slides]
19.12.2016	paper presentations	(Atilla) On subjective measures of interestingness in knowledge discovery Silberschatz & Tuzhilin, KDD 1995 (Arthur) Random walk inference and learning in a large scale knowledge base Lao et al., EMNLP 2011 [slides]
9.01.2017	paper presentations	(Anna) Authorship verification as a one-class classification problem Koppel & Schler, ICML 2004 [slides] (Irina) Learning to explain entity relationships in knowledge graphs Voskarides et al., ACL-IJCNLP 2015 [slides]
16.01.2017	paper presentations	(Zoia) Knowledge base completion via search-based question answering West et al., WWW 2014 [slides] (Daaje) Discovering factions in computational linguistics community Sim et al., ACL Workshop on rediscovering 50 years of discoveries, 2012 [slides]
23.01.2017	lecture	Introduction to topic models: part I -- the big picture and refresher on probabilities [slides]
30.01.2017	lecture	Introduction to topic models: part II -- the mathematical story [slides]
6.02.2017	project ideas presentations	10 minute presentations for the homework projects.
14.04.2017	projects due!	Together with the code I expect a project report, organized similarly to a paper: have an introduction that explains your idea related work to show on which previous work you base your idea and your implementation describe the data you are using, the experimental set-up, the experiments and results and very importantly DISCUSSION (!!!) And some conclusions about what you learned during the project and what the experiments tell about your initial idea. There is no minimum or maximum limit on the length of this document, but give enough details to convince me you did a good job while sparing me irrelevant details.

Literatur

Not defined yet