Interactive Fiction Game Playing as Multi-Paragraph Reading Comprehension with Reinforcement Learning
By Xiaoxiao Guo, Mo Yu, Yupeng Gao, Chuang Gan, Murray Campbell, Shiyu Chang. 2020
35 citations recorded by the source corpus.
Interactive Fiction (IF) games with real human-written natural language texts provide a new natural evaluation for language understanding techniques. In contrast to previous text games with mostly synthetic texts, IF games pose language understanding challenges on the human-written textual descriptions of diverse and sophisticated game worlds and language generation challenges on the action command generation from less restricted combinatorial space. We take a novel perspective of IF game solving and re-formulate it as Multi-Passage Reading Comprehension (MPRC) tasks. Our approaches utilize the context-query attention mechanisms and the structured prediction in MPRC to efficiently generate and evaluate action outputs and apply an object-centric historical observation retrieval strategy to mitigate the partial observability of the textual observations. Extensive experiments on the recent IF benchmark (Jericho) demonstrate clear advantages of our approaches achieving high winning rates and low data requirements compared to all previous approaches. Our source code is available at: this https URL.
Subfields: Artificial Intelligence, Developmental and Educational Psychology, Natural Language Processing, Reinforcement Learning, Question Answering & Reading Comprehension
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Shunyu Yao
- Karthik Narasimhan
- Tom Young
- Matthew J. Hausknecht
- Erik Cambria
- Jinjie Ni
- Vlad Pandelea
- Fuzhao Xue
In Natural Language Processing
- Shunyu Yao
- Karthik Narasimhan
- Matthew J. Hausknecht
- Jinjie Ni
- Vlad Pandelea
- Tom Young
- Fuzhao Xue
- Vinay Vishnumurthy Adiga
In Reinforcement Learning
- Shunyu Yao
- Karthik Narasimhan
- Matthew J. Hausknecht
- Jinjie Ni
- Vlad Pandelea
- Tom Young
- Fuzhao Xue
- Vinay Vishnumurthy Adiga
In Question Answering & Reading Comprehension
Citations in the corpus
- Multi-Stage Episodic Control for Strategic Exploration in Text Games (2022, unknown): “The policy π exploit is updated every n episodes using the process in Section 3.2, while π explore is updated within episodes at every step using the TD loss in equation 7, sampling high rewarding trajectories with priority fraction ρ , similar to (Guo et al., 2020) 5 .”
- Reading and Acting while Blindfolded: The Need for Semantics in Text Game Agents (2021, background): “, 2020), and reading comprehension systems (Guo et al., 2020).”
- Recent advances in deep learning based dialogue systems: a systematic survey (2021, background): “Dialogue-related tasks like dialogue relation extraction (Li et al., 2019b), question answering (Hua et al., 2020) and machine reading comprehension (Guo et al., 2020) benefit from reinforcement learning as well because of their interactive nature and the scarcity of annotated data.”