Knowledge Extraction from Games Task
A Knowledge Extraction from Games Task is a knowledge extraction task from game playing.
- Example(s):
- Contextual query-answering in games where non-player characters (or visual cues in environment design) offer hints to solve problems
- Extracting architectural information from game level layouts
- Transfer learning, analogical reasoning, or goal reasoning within or between games or game levels
- Game-playing agents which can explain their own actions or policy in terms of the game’s rules
- Learning the rules of a game from observation, or learning higher-level rules or goals automatically
- Determining a designer or player’s mental model of game rules, and whether that differs from the rules induced by the game’s implementation
- See: Template Filling (IE) Task.
References
2017
- http://www.digra.org/cfp-aaai-workshop-on-knowledge-extraction-from-games/
- QUOTE: Knowledge Extraction from Games (KEG) is a new workshop exploring questions of and approaches to the mechanical extraction of *knowledge* from games meant for humans — including but not limited to game rules, character graphics, environment maps, music and sound effects, high-level goals or heuristic strategies, transferrable skills, aesthetic standards and conventions, or abstracted models of games.
It includes and expands on the mandate of a recent vision paper at Computational Intelligence in Games (CIG 2017), [Automated Game Design Learning](http://www.cig2017.com/wp-content/uploads/2017/08/paper_95.pdf).
Games provide useful structuring information for many reasoning tasks and therefore provide interesting environments for knowledge extraction and specification recovery.
Some examples of work that would be appropriate for KEG include:
- Contextual query-answering in games where non-player characters (or visual cues in environment design) offer hints to solve problems
- Extracting architectural information from game level layouts
- Transfer learning, analogical reasoning, or goal reasoning within or between games or game levels
- Game-playing agents which can explain their own actions or policy in terms of the game’s rules
- Learning the rules of a game from observation, or learning higher-level rules or goals automatically
- Determining a designer or player’s mental model of game rules, and whether that differs from the rules induced by the game’s implementation
- QUOTE: Knowledge Extraction from Games (KEG) is a new workshop exploring questions of and approaches to the mechanical extraction of *knowledge* from games meant for humans — including but not limited to game rules, character graphics, environment maps, music and sound effects, high-level goals or heuristic strategies, transferrable skills, aesthetic standards and conventions, or abstracted models of games.