Please use this identifier to cite or link to this item: http://hdl.handle.net/1893/27857
Appears in Collections:Computing Science and Mathematics Book Chapters and Sections
Title: A Case Study of Closed-Domain Response Suggestion with Limited Training Data
Author(s): Galke, Lukas
Gerstenkorn, Gunnar
Scherp, Ansgar
Editor(s): Elloumi, M
Granitzer, M
Hameurlain, A
Seifert, C
Stein, B
Tjoa, AM
Wagner, R
Sponsor: European Commission
Citation: Galke L, Gerstenkorn G & Scherp A (2018) A Case Study of Closed-Domain Response Suggestion with Limited Training Data. In: Elloumi M, Granitzer M, Hameurlain A, Seifert C, Stein B, Tjoa A & Wagner R (eds.) Communications in Computer and Information Science; Database and Expert Systems Applications. Communications in Computer and Information Science, 903. Regensburg, Germany: Springer International Publishing, pp. 218-229. https://doi.org/10.1007/978-3-319-99133-7_18
Issue Date: 31-Dec-2018
Series/Report no.: Communications in Computer and Information Science, 903
Abstract: We analyze the problem of response suggestion in a closed domain along a real-world scenario of a digital library. We present a text-processing pipeline to generate question-answer pairs from chat transcripts. On this limited amount of training data, we compare retrieval-based, conditioned-generation, and dedicated representation learning approaches for response suggestion. Our results show that retrieval-based methods that strive to find similar, known contexts are preferable over parametric approaches from the conditioned-generation family, when the training data is limited. We, however, identify a specific representation learning approach that is competitive to the retrieval-based approaches despite the training data limitation.
Rights: This is a post-peer-review, pre-copyedit version of a paper published in Elloumi M, Granitzer M, Hameurlain A, Seifert C, Stein B, Tjoa A & Wagner R (eds.) Database and Expert Systems Applications. DEXA 2018. Communications in Computer and Information Science, 903. The final authenticated version is available online at: https://doi.org/10.1007/978-3-319-99133-7_18
DOI Link: 10.1007/978-3-319-99133-7_18

Files in This Item:
File Description SizeFormat 
W42-GalkeEtAl-A Case Study of Closed-Domain Response Suggestion with Limited Training Data.pdfFulltext - Accepted Version255.49 kBAdobe PDFView/Open



This item is protected by original copyright



Items in the Repository are protected by copyright, with all rights reserved, unless otherwise indicated.

If you believe that any material held in STORRE infringes copyright, please contact library@stir.ac.uk providing details and we will remove the Work from public display in STORRE and investigate your claim.