Mode of access: World Wide Web.

Part of: Synthesis digital library of engineering and computer science.

Includes bibliographical references (pages 133-139).

1. Introduction -- 1.1 Evolutionary computation -- 1.2 Elements of fitness for level design -- 1.3 Obscured mazes: a simple example -- 1.3.1 Chess mazes -- 1.3.2 Chromatic mazes -- 1.3.3 Key maps and alternate views -- 1.4 Conclusions --

2. Contrasting representations for maze generation -- 2.1 Details of the binary direct representation -- 2.2 Details of the chromatic representation -- 2.3 Details of the positive, indirect representation -- 2.4 Details of the negative, indirect representation -- 2.5 Fitness function design -- 2.5.1 Definitions -- 2.5.2 Fitness functions -- 2.6 Design of experiments -- 2.6.1 Initial experiments -- 2.6.2 Experiments with culs-de-sac -- 2.6.3 Changing the board size -- 2.6.4 Experiments with the chromatic representation -- 2.6.5 Verification of sparse initialization and crossover -- 2.7 Results and discussion for maze generation -- 2.7.1 Experiments with culs-de-sac -- 2.7.2 Experiments with different board sizes -- 2.7.3 Sparse initialization and choice of crossover operator -- 2.7.4 Algorithm speed -- 2.7.5 Fitness landscapes and sparse initialization -- 2.7.6 Discussion for maze generation -- 2.7.7 Breaking out of two dimensions --

3. Dual mazes -- 3.1 Representations for dual maze generation -- 3.2 Details of the generative representation -- 3.2.1 Details of the direct representation -- 3.2.2 Fitness function specification -- 3.3 Experimental design -- 3.4 Results and discussion for dual mazes -- 3.5 Conclusions and next steps for dual mazes -- 3.5.1 Additional fitness elements -- 3.5.2 Tool development -- 3.5.3 Visibility and lines of sight -- 3.5.4 Terrain types --

4. Terrain maps -- 4.1 Midpoint L-systems -- 4.1.1 The representation for midpoint L-systems -- 4.1.2 Multiscale landforms -- 4.2 Landscape automata: another representation for height maps -- 4.2.1 Defining landscape automata -- 4.2.2 Experiments with landscape automata -- 4.2.3 Results and discussion for landscape automata -- 4.2.4 Qualitative diversity -- 4.2.5 Conclusions and next steps for landscape automata -- 4.3 Morphing and smoothing of height maps --

5. Cellular automata based maps -- 5.1 Fashion-based cellular automata -- 5.1.1 Design of experiments -- 5.1.2 Results and discussion for cellular automata level creation -- 5.1.3 Discussion for cellular automata level design -- 5.1.4 Using an optimizer for non-optimization goals -- 5.2 Generalizing fitness and morphing -- 5.2.1 Generalizing the fitness function to control open space -- 5.2.2 Return of dynamic programming based fitness -- 5.2.3 Morphing between rules -- 5.2.4 More general application of morphing: re-evolution --

6. Decomposition, tiling, and assembly -- 6.1 More maps than you could ever use -- 6.1.1 Details of tile production -- 6.1.2 Enumerating maps and exploiting tile symmetries -- 6.2 Required content -- 6.2.1 The fitness function for required content tiles -- 6.2.2 Results of the tile creation experiments -- 6.3 Creating an integrated adventure: goblins attack the village -- 6.3.1 System design for FRPG module creation -- 6.3.2 The level evolver -- 6.3.3 Identifying and connecting rooms -- 6.3.4 Populating the dungeon -- 6.3.5 Results for FRPG module creation -- 6.3.6 Conclusions and next steps for FRPG module generation -- 6.3.7 Decorations: monsters, treasure, and traps -- 6.3.8 History, context, and story --

Bibliography -- Author's biography.

Abstract freely available; full-text restricted to subscribers or individual document purchasers.

Compendex

INSPEC

Google scholar

Google book search

Automatic content generation is the production of content for games, web pages, or other purposes by procedural means. Search-based automatic content generation employs search-based algorithms to accomplish automatic content generation. This book presents a number of different techniques for search-based automatic content generation where the search algorithm is an evolutionary algorithm. The chapters treat puzzle design, the creation of small maps or mazes, the use of L-systems and a generalization of L-system to create terrain maps, the use of cellular automata to create maps, and, finally, the decomposition of the design problem for large, complex maps culminating in the creation of a map for a fantasy game module with designer-supplied content and tactical features. The evolutionary algorithms used for the different types of content are generic and similar, with the exception of the novel sparse initialization technique are presented in Chapter 2. The points where the content generation systems vary are in the design of their fitness functions and in the way the space of objects being searched is represented. A large variety of different fitness functions are designed and explained, and similarly radically different representations are applied to the design of digital objects all of which are, essentially, maps for use in games.

Also available in print.

Title from PDF title page (viewed on May 25, 2018).