KIT | KIT-Bibliothek | Impressum | Datenschutz

What’s the Matter? Knowledge Acquisition by Unsupervised Multi-Topic Labeling for Spoken Utterances

Weigelt, Sebastian; Keim, Jan; Hey, Tobias ORCID iD icon; Tichy, Walter F. ORCID iD icon

Abstract (englisch):

Systems such as Alexa, Cortana, and Siri app ear rather smart. However, they only react to predefined wordings and do not actually grasp the user's intent. To overcome this limitation, a system must understand the topics the user is talking about. Therefore, we apply unsupervised multi-topic labeling to spoken utterances. Although topic labeling is a well-studied task on textual documents, its potential for spoken input is almost unexplored. Our approach for topic labeling is tailored to spoken utterances; it copes with short and ungrammatical input.
The approach is two-tiered. First, we disambiguate word senses. We utilize Wikipedia as pre-labeled corpus to train a naïve-bayes classifier. Second, we build topic graphs based on DBpedia relations. We use two strategies to determine central terms in the graphs, i.e. the shared topics. One fo cuses on the dominant senses in the utterance and the other covers as many distinct senses as possible. Our approach creates multiple distinct topics per utterance and ranks results.
The evaluation shows that the approach is feasible; the word sense disambiguation achieves a recall of 0.799. Concerning topic labeling, in a user study subjects assessed that in 90.9% of the cases at least one proposed topic label among the first four is a good fit. ... mehr


Verlagsausgabe §
DOI: 10.5445/IR/1000140308
Veröffentlicht am 29.11.2021
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Programmstrukturen und Datenorganisation (IPD)
Publikationstyp Zeitschriftenaufsatz
Publikationsmonat/-jahr 08.2020
Sprache Englisch
Identifikator ISSN: 2641-953X
KITopen-ID: 1000140308
Erschienen in International Journal of Humanized Computing and Communication
Verlag Institute for Semantic Computing Foundation
Band 1
Heft 1
Seiten 43–66
Vorab online veröffentlicht am 01.08.2020
Schlagwörter Topic Labeling, Topic Modeling, Unsupervised Machine Learning, Graph Centrality Measures, Word Sense Disambiguation, Ontology Selection, DBpedia, Wikipedia, Semantic Annotation, Spoken Language Interfaces, Spoken Language Understanding, Natural Language Processing
Nachgewiesen in Dimensions
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page