Seminar in summer term 2021: NLP for Software Engineering

TU München - Fakultät für Informatik
Software- and Systems Engineering Research Group

Seminar in the summer term 2021
NLP for Software Engineering

Seminare

Inhalt

Natural Language Processing (NLP) deals with the automatic analysis of natural language. Applications are for example the extraction of information from large amounts of text, automatic translation or automatic dialog systems. In recent years, the field has received a further boost through the emergence of deep-learning techniques, whereby numerous NLP tasks can now be solved with high quality. But also classical, proven NLP techniques are still in use.

In this seminar we want to look at some fields of application for NLP. The focus of the seminar is less the basics of NLP but more their application. Based on concrete data and tasks NLP techniques will be tried out and the result will be evaluated. In the seminar thisis and the seminar lecture the techniques as well as the result of the application will be presented. In addition to general use cases, possibilities of how NLP techniques can be used in the context of software or requirements engineering will be investigated.

News

The seminar will take place in the summer term 2021.
The seminar briefing will take place on February 2nd, 10am-11am via Zoom (for details see TUMOnline)

Dates

2021/02/02 10am-11am: Seminar briefing via Zoom (see TUMOnline for details)

Topics

Text-Classification and Extraction

Sentiment Analysis (Objectivity Analyse)
Question Answering / Information Extraction
Topic Extraction

Text Transformation

Maschine Translation
Automatic Summarization

Detecting References & Relationships

Coreference Analysis
Extraction of Object References

Application to Software-Requirements

Detecting Semantic Duplicates
Cause-Effect Extraction

Learning objectives

Each participant learns to work independently on a topic, to conduct literature research, to apply the knowledge gained and to write a paper on the given topic. Each participant gives a presentation on his or her topic in front of the other participants and supervisors and receives detailed feedback.

Success Criteria

Attendance at all appointments, preparation of own seminar paper, application of knowledge in a given case study and presentation.

Prerequisites

Basics in NLP / Machine Learning / Deep Learning (e.g. Natural Language Processing IN2361). Good (at least passive) knowledge of English, since the required literature is usually in English.

Registration

Registration details will be announced after the seminar briefing. Those interested in the seminar should send an application (status in studies, interest in software quality, motivation, which topic is desired?) to junkerm@in.tum.de by february, 11th.

Organizer

The seminar is organized by