Distinctive Slogan Generation with Reconstruction
Shotaro Misawa, Yasuhide Miura, Tomoki Taniguchi, Tomoko Ohkuma
Abstract
E-commerce sites include advertising slogans along with information regarding an item. Slogans can attract viewers’ attention to increase sales or visits by emphasizing advantages of an item. The aim of this study is to generate a slogan from a description of an item. To generate a slogan, we apply an encoder–decoder model which has shown effectiveness in many kinds of natural language generation tasks, such as abstractive summarization. However, slogan generation task has three characteristics that distinguish it from other natural language generation tasks: distinctiveness, topic emphasis, and style difference. To handle these three characteristics, we propose a compressed representation–based reconstruction model with refer–attention and conversion layers. The results of the experiments indicate that, based on automatic and human evaluation, our method achieves higher performance than conventional methods.- Anthology ID:
- 2020.ecomnlp-1.9
- Volume:
- Proceedings of Workshop on Natural Language Processing in E-Commerce
- Month:
- Dec
- Year:
- 2020
- Address:
- Barcelona, Spain
- Editors:
- Huasha Zhao, Parikshit Sondhi, Nguyen Bach, Sanjika Hewavitharana, Yifan He, Luo Si, Heng Ji
- Venue:
- EcomNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 87–97
- Language:
- URL:
- https://aclanthology.org/2020.ecomnlp-1.9
- DOI:
- Cite (ACL):
- Shotaro Misawa, Yasuhide Miura, Tomoki Taniguchi, and Tomoko Ohkuma. 2020. Distinctive Slogan Generation with Reconstruction. In Proceedings of Workshop on Natural Language Processing in E-Commerce, pages 87–97, Barcelona, Spain. Association for Computational Linguistics.
- Cite (Informal):
- Distinctive Slogan Generation with Reconstruction (Misawa et al., EcomNLP 2020)
- PDF:
- https://preview.aclanthology.org/proper-vol2-ingestion/2020.ecomnlp-1.9.pdf