Automated Generation of Conversational Non Player Characters

2017 IEEE Conference on Computational Intelligence and Games (CIG), 2017

We present and describe CiF-CK-a social agent architecture that models reasoning about persistent social interactions to improve narrative engagement and play experience for human interactors. The architecture is inspired by McCoy et al's Comme il-Faut (CiF) architecture that represented rich social interactions between agents that included feelings, social and relationship contexts, and longer term mood. The key contribution of this work is in adapting the richness of social interactions from CiF to a first-person interaction experience and a released distribution of its implementation on the Skyrim game engine. The released modification has been successful in the player community for the popular game.

This is a thesis that deals with the norms and rules of players playing online games together. It is also a thesis about believability, the current capabilities of non-player characters (NPCs) and the attitudes amongst game developers towards dynamic and systemic games AI.

Loading..., 2009

This paper investigates issues of character believability and intelligent personalization through a reading of the Elder Scrolls: Oblivion. Oblivion's opening sequence simultaneously trains players in the function of the game, and allows them to customize their character class through the choices and actions they take. Oblivion makes an ambitious attempt at intelligent personalization in the character creation process. Its strategy is to track early gameplay decisions and "stereotype" players into one of 21 possible classes. This approach has two advantages over a less adaptive system. First, it supports the illusion of the gameworld as a real world by embedding the process of character creation within a narrativised gameplay context. Second, the intelligent recommendation system responds to the player's desire to believe that the game "knows" something about her personality. This leads the players to conceptualize the system as an entity with autonomous, human-like knowledge. This paper considers ways in which Oblivion both succeeds and fails at mapping player behaviour to appropriate class assignments. It does so through the analysis of multiple replayings of the opening sequence, and the application of two theoretical lenses-character believability and intelligent personalization. The paper documents moments where the dialogue between player and game breaks down, and argues for alternative techniques to customize the play experience within the desires of the player.

2008

Figure 1. Sample animated heads showing expressions, talking, and moving, all created from one synergistic system.

Proc. 3rd Digital Games Research Association International Conference, 2007

Descriptions of humans require several qualities for people to experience them as believable: human body; selfawareness, intentional states, and self impelled actions; expression of emotions; ability to use natural language; and persistent traits. Based on these we analyze non-player character Claudette Perrick in The Elders Scroll IV:

2010

Computing technologies have advanced rapidly over the past decade. Faster machines, better graphics, and more advanced algorithms become available every year. Moreover, the evolution of internet technology and the increasing accessibility of computing resources and mobile devices allow computing technologies to go beyond business and scientific computing, and become an important means for providing entertainment and facilitating communication.

2008

Recent years have seen a growing interest in creating virtual agents to populate the cast of characters for interactive narrative. A key challenge posed by interactive characters for narrative environments is devising expressive dialogue generators. To be effective, character dialogue generators must be able to simultaneously take into account multiple sources of information that bear on dialogue, including character attributes, plot development, and communicative goals.

NDUB Research Journal, 2023

Character dialogue writing for current digital games is a challenging task because many of them are non-linear. It requires authors to produce many variations of the characters' lines in order to make the game appearance more genuine. The proposed approach is to replace game cut-scenes with voice interactive conversation, which uses a speech recognition technology to speak with NPCs (Non-player charactes) in natural language like English. This is part of the game development field. The conversation is point-and-click-based, non-interactive but it does not handle mood, according to previous systems investigated. Designing an algorithm that allows NPCs to reply to natural language is the solution here. This allows vocal communication with NPCs, which increases realism in games. In this work, we design an algorithm that allows for spoken communication with the NPC. The goal of this work is to make engaging with NPCs, who speak to players in normal English, straightforward, realistic, and quick. After qualitative and quantitative testing using survey on some players, the results show that it is possible to use the algorithm during game development or patch it after development. We check how NPCs respond seamlessly to the natural language. We also make sure that the majority of the players' vocal interactions made the game more realistic and highly immersive. And we confirm that it is a completely seamless integration during development. Finally, we ensure that the player's opinion about increasing immersion has made the game more enjoyable than the vanilla version.

2014