Abstract
Following navigation instructions in natural language requires a composition of language, action, and knowledge of the environment. Knowledge of the environment may be provided via visual sensors or as a symbolic world representation referred to as a map. Here we introduce the Realistic Urban Navigation (RUN) task, aimed at interpreting navigation instructions based on a real, dense, urban map. Using Amazon Mechanical Turk, we collected a dataset of 2515 instructions aligned with actual routes over three regions of Manhattan. We propose a strong baseline for the task and empirically investigate which aspects of the neural architecture are important for the RUN success. Our results empirically show that entity abstraction, attention over words and worlds, and a constantly updating world-state, significantly contribute to task accuracy.
| Original language | English |
|---|---|
| Title of host publication | EMNLP-IJCNLP 2019 - 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing, Proceedings of the Conference |
| Publisher | Association for Computational Linguistics |
| Pages | 6449-6455 |
| Number of pages | 7 |
| ISBN (Electronic) | 9781950737901 |
| State | Published - 2019 |
| Externally published | Yes |
| Event | 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing, EMNLP-IJCNLP 2019 - Hong Kong, China Duration: 3 Nov 2019 → 7 Nov 2019 |
Publication series
| Name | EMNLP-IJCNLP 2019 - 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing, Proceedings of the Conference |
|---|
Conference
| Conference | 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing, EMNLP-IJCNLP 2019 |
|---|---|
| Country/Territory | China |
| City | Hong Kong |
| Period | 3/11/19 → 7/11/19 |
Bibliographical note
Publisher Copyright:© 2019 Association for Computational Linguistics
Funding
We thank Yoav Goldberg and Yoav Artzi, for their advice and comments. We thank the ONLP team at the Open University of Israel, for fruitful discussions throughout the process. We further thank the anonymous reviewers for their helpful comments. This research was presented at Georgetown University, Cornell University, and Tel-Aviv University. This research is supported by European Research Council, ERC-StG-2015 scheme, Grant number 677352, and by the Israel Science Foundation (ISF), Grant number 1739/26, for which we are grateful.
| Funders | Funder number |
|---|---|
| Open University of Israel | |
| Tel-Aviv University | |
| Cornell University | |
| Georgetown University | |
| Horizon 2020 Framework Programme | 677352 |
| European Commission | ERC-StG-2015 |
| Israel Science Foundation | 1739/26 |