Please use this identifier to cite or link to this item:
https://doi.org/10.21256/zhaw-4889
Publication type: | Conference paper |
Type of review: | Not specified |
Title: | End-to-end trainable system for enhancing diversity in natural language generation |
Authors: | Deriu, Jan Milan Cieliebak, Mark |
DOI: | 10.21256/zhaw-4889 |
Conference details: | End-to-End Natural Language Generation Challenge (E2E NLG), 2017 |
Issue Date: | 2017 |
Publisher / Ed. Institution: | ZHAW Zürcher Hochschule für Angewandte Wissenschaften |
Language: | English |
Subject (DDC): | 006: Special computer methods |
Abstract: | Natural Language Generation plays an important role in the domain of dialogue systems as it determines how the users perceive the system. Recently, deep-learning based systems have been proposed to tackle this task, as they generalize better and do not require large amounts of manual effort to implement them for new domains. However, deep learning systems usually produce monotonous sounding texts. In this work, we present our system for Natural Language Generation where we control the first word of the surface realization. We show that with this simple control mechanism it is possible to increase the lexical variability and the complexity of the generated texts. For this, we apply a character-based version of the Semantically Controlled Long Short-term Memory Network (SC-LSTM), and apply its specialized cell to control the first word generated by the system. To ensure that the surface manipulation does not produce semantically incoherent texts we apply a semantic control component, which we also use for reranking purposes. We show that our model is capable of generating texts that are more sophisticated while decreasing the number of semantic errors made during the generation. |
URI: | http://www.macs.hw.ac.uk/InteractionLab/E2E/final_papers/E2E-ZHAW.pdf https://digitalcollection.zhaw.ch/handle/11475/13220 |
Fulltext version: | Published version |
License (according to publishing contract): | Licence according to publishing contract |
Departement: | School of Engineering |
Organisational Unit: | Institute of Computer Science (InIT) |
Appears in collections: | Publikationen School of Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
2017_Deriu_End_to_end_trainable_system_for_enhancing_diversity.pdf | 433.76 kB | Adobe PDF | View/Open |
Show full item record
Deriu, J. M., & Cieliebak, M. (2017). End-to-end trainable system for enhancing diversity in natural language generation. End-to-End Natural Language Generation Challenge (E2E NLG), 2017. https://doi.org/10.21256/zhaw-4889
Deriu, J.M. and Cieliebak, M. (2017) ‘End-to-end trainable system for enhancing diversity in natural language generation’, in End-to-End Natural Language Generation Challenge (E2E NLG), 2017. ZHAW Zürcher Hochschule für Angewandte Wissenschaften. Available at: https://doi.org/10.21256/zhaw-4889.
J. M. Deriu and M. Cieliebak, “End-to-end trainable system for enhancing diversity in natural language generation,” in End-to-End Natural Language Generation Challenge (E2E NLG), 2017, 2017. doi: 10.21256/zhaw-4889.
DERIU, Jan Milan und Mark CIELIEBAK, 2017. End-to-end trainable system for enhancing diversity in natural language generation. In: End-to-End Natural Language Generation Challenge (E2E NLG), 2017 [online]. Conference paper. ZHAW Zürcher Hochschule für Angewandte Wissenschaften. 2017. Verfügbar unter: http://www.macs.hw.ac.uk/InteractionLab/E2E/final_papers/E2E-ZHAW.pdf
Deriu, Jan Milan, and Mark Cieliebak. 2017. “End-to-End Trainable System for Enhancing Diversity in Natural Language Generation.” Conference paper. In End-to-End Natural Language Generation Challenge (E2E NLG), 2017. ZHAW Zürcher Hochschule für Angewandte Wissenschaften. https://doi.org/10.21256/zhaw-4889.
Deriu, Jan Milan, and Mark Cieliebak. “End-to-End Trainable System for Enhancing Diversity in Natural Language Generation.” End-to-End Natural Language Generation Challenge (E2E NLG), 2017, ZHAW Zürcher Hochschule für Angewandte Wissenschaften, 2017, https://doi.org/10.21256/zhaw-4889.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.