Beyond Utterance Extraction: Summary Recombination for Speech Summarization

Trione, Jérémy; Favre, Benoît; Béchet, Frédéric

doi:10.21437/interspeech.2016-855

Cited by 2 publications

(1 citation statement)

References 9 publications

(8 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The main contribution is the proposal of a French language corpus originated from the Parisian public transportation call center, which ought to reduce the development cost of speech analytics systems by limiting the need for manual data annotation. Based on such a corpus, several studies have been conducted, including theme identification (Estève et al, 2015;Morchid et al, 2013Morchid et al, , 2014aParcollet et al, 2018), dialogue classification (Koço et al, 2012;Morchid et al, 2014b), named entity and semantic concept extraction (Ghannay et al, 2018), and speech summarization (Trione et al, 2016).…”

Section: An Account Of Related Workmentioning

confidence: 99%

A combined approach to the analysis of speech conversations in a contact center domain

Brunello¹,

Marzano²,

Montanari³

et al. 2022

Preprint

View full text Add to dashboard Cite

The ever more accurate search for deep analysis in customer data is a really strong technological trend nowadays, quite appealing to both private and public companies. This is particularly true in the contact center domain, where speech analytics is an extremely powerful methodology for gaining insights from unstructured data, coming from customer and human agent conversations. In this work, we describe an experimentation with a speech analytics process for an Italian contact center, that deals with call recordings extracted from inbound or outbound flows. First, we illustrate in detail the development of an in-house speech-to-text solution, based on Kaldi framework, and evaluate its performance (and compare it to Google Cloud Speech API). Then, we evaluate and compare different approaches to the semantic tagging of call transcripts, ranging from classic regular expressions to machine learning models based on ngrams and logistic regression, and propose a combination of them, which is shown to provide a consistent benefit. Finally, a decision tree inducer, called J48S, is applied to the problem of tagging. Such an algorithm is natively capable of exploiting sequential data, such as texts, for classification purposes. The solution is compared with the other approaches and is shown to provide competitive classification performances, while generating highly interpretable models and reducing the complexity of the data preparation phase. The potential operational impact of the whole process is thoroughly examined.

show abstract