Generation of game contents by social media analysis and MAS planning

In the age of pervasive computing and social networks, it has become commonplace to retrieve opinions about digital contents in games. In the case of multi-player, open world gaming, in fact even in “old-school” single players games, it is evident the need for adding new features in a game depending on users comments and needs. However this is a challenging task that usually requires considerable design and programming efforts, and more and more patches to games, with the inevitable consequence of loosing interest in the game by players over years. This is particularly a hard problem for all games that do not intend to be designed as interactive novels. Process Content Generation (PCG) of new contents could be a solution to this problem, but usually such techniques are used to design new maps or graphical contents. Here we propose a novel PCG technique able to introduce new contents in games by means of new story-lines and quests. We introduce new intelligent agents and events in the world: their attitudes and behaviors will promote new actions in the game, leading to the involvement of players in new gaming content. The whole methodology is driven by Social Media Analysis contents about the game, and by the use of formal planning techniques based on Multi-Agents models. © 2019 Elsevier Lt