Baselines for Reinforcement Learning in Text Games

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F18%3A10387543" target="_blank" >RIV/00216208:11320/18:10387543 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.1109/ICTAI.2018.00058" target="_blank" >http://dx.doi.org/10.1109/ICTAI.2018.00058</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICTAI.2018.00058" target="_blank" >10.1109/ICTAI.2018.00058</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Baselines for Reinforcement Learning in Text Games

Popis výsledku v původním jazyce

The ability to learn optimal control policies in systems where action space is defined by sentences in natural language would allow many interesting real-world applications such as automatic optimisation of dialogue systems. Text-based games with multiple endings and rewards are a promising platform for this task, since their feedback allows us to employ reinforcement learning techniques to jointly learn text representations and control policies. We argue that the key property of AI agents, especially in the text-games context, is their ability to generalise to previously unseen games. We present a minimalistic text-game playing agent, testing its generalisation and transfer learning performance and showing its ability to play multiple games at once. We also present pyfiction, an open-source library for universal access to different text games that could, together with our agent that implements its interface, serve as a baseline for future research.

Název v anglickém jazyce

Baselines for Reinforcement Learning in Text Games

Popis výsledku anglicky

The ability to learn optimal control policies in systems where action space is defined by sentences in natural language would allow many interesting real-world applications such as automatic optimisation of dialogue systems. Text-based games with multiple endings and rewards are a promising platform for this task, since their feedback allows us to employ reinforcement learning techniques to jointly learn text representations and control policies. We argue that the key property of AI agents, especially in the text-games context, is their ability to generalise to previously unseen games. We present a minimalistic text-game playing agent, testing its generalisation and transfer learning performance and showing its ability to play multiple games at once. We also present pyfiction, an open-source library for universal access to different text games that could, together with our agent that implements its interface, serve as a baseline for future research.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/GJ17-17125Y" target="_blank" >GJ17-17125Y: Balancování deliberativního a reaktivního chování inteligentních agentů</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2018

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of ICTAI 2018 : International Conference on Tools with Artificial Intelligence

ISBN

978-1-5386-7449-9

ISSN

1082-3409

e-ISSN

neuvedeno

Počet stran výsledku

8

Strana od-do

320-327

Název nakladatele

IEEE

Místo vydání

Volos, Greece

Místo konání akce

Volos,Greece

Datum konání akce

5. 11. 2018

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000457750200048