Goto

Collaborating Authors

New York University


De Mesentier Silva

AAAI Conferences

The process of play testing a game is subjective, expensive and incomplete. In this paper, we present a play-testing approach that explores the game space with automated agents and collects data to answer questions posed by the designers. Rather than have agents interacting with an actual game client, this approach recreates the bare bone mechanics of the game as a separate system. Our agent is able to play in minutes what would take testers days of organic gameplay. The analysis of thousands of game simulations exposed imbalances in game actions, identified inconsequential rewards and evaluated the effectiveness of optional strategic choices. Our test case game, The Sims Mobile, was recently released and the findings shown here influenced design changes that resulted in improved player experience.


Exploring Gameplay With AI Agents

AAAI Conferences

The process of play testing a game is subjective, expensive and incomplete. In this paper, we present a play-testing approach that explores the game space with automated agents and collects data to answer questions posed by the designers. Rather than have agents interacting with an actual game client, this approach recreates the bare bone mechanics of the game as a separate system. Our agent is able to play in minutes what would take testers days of organic gameplay. The analysis of thousands of game simulations exposed imbalances in game actions, identified inconsequential rewards and evaluated the effectiveness of optional strategic choices. Our test case game, The Sims Mobile, was recently released and the findings shown here influenced design changes that resulted in improved player experience.


Games as Co-Creative Cooperative Systems

AAAI Conferences

Many modern creative industrial processes rely on the collaboration between multiple humans, assisted by one or more computational systems, in a complex environment. However, most traditional systems lack the adaptability required to contribute in a flexible, co-creative manner, instead executing a fixed set of tasks in a preset time schedule. We believe games, especially cooperative games offer an ideal platform to conduct research in co-creativity. We present our motivation, preliminary work and future goals to study, build and measure game-inspired co-creative AI systems.


Like a DNA String: Sequence-Based Player Profiling in Tom Clancy’s The Division

AAAI Conferences

In this paper we present an approach to using sequence analysis to model player behavior. This approach is designed to work in game development contexts, integrating production teams and delivering profiles that inform game design. We demonstrate the method via a case study of the game T om Clancy’s The Division, which with its 20 million players represents a major current commercial title. The approach presented provides a mixed-methods framework, combining qualitative knowledge elicitation and workshops with large-scale telemetry analysis, using sequence mining and clustering to develop detailed player profiles showing the core game-play loops of The Division’s players.


From the User to the Medium: Neural Profiling Across Web Communities

AAAI Conferences

Online communities provide a unique way for individuals to access information from those in similar circumstances, which can be critical for health conditions that require daily and personalized management. As these groups and topics often arise organically, identifying the types of topics discussed is necessary to understand their needs. As well, these communities and people in them can be quite diverse, and existing community detection methods have not been extended towards evaluating these heterogeneities. This has been limited as community detection methodologies have not focused on community detection based on semantic relations between textual features of the user-generated content. Thus here we develop an approach, NeuroCom, that optimally finds dense groups of users as communities in a latent space inferred by neural representation of published contents of users. By embedding of words and messages, we show that NeuroCom demonstrates improved clustering and identifies more nuanced discussion topics in contrast to other common unsupervised learning approaches.


On Group Popularity Prediction in Event-Based Social Networks

AAAI Conferences

Although previous work has shown that member and structural features are important to the future popularity of groups in EBSN, it is not yet clear how different member roles and the interplay between them contribute to group popularity. In this paper, we study a real-world dataset from Meetup --- a popular EBSN platform --- and propose a deep neural network based method to predict the popularity of new Meetup groups. Our method uses group-level features specific to event-based social networks, such as time and location of events in a group, as well as the structural features internal to a group, such as the inferred member roles in a group and social substructures among members. Empirically, our approach reduces the RMSE of the popularity prediction (measured in RSVPs) of a group's future events by up to 12%, against the state-of-the-art baselines.


Khatibi

AAAI Conferences

Accurate predictions about future events is essential in many areas, one of them being the Tourism Industry. Usually, countries and cities invest a huge amount of money in planning and preparation in order to welcome (and profit from) tourists. An accurate prediction of the number of visits in the following days or months could help both the economy and tourists. Prior studies in this domain explore forecasting for a whole country rather than for fine-grained areas within a country (e.g., specific touristic attractions). In this work, we suggest that accessible data from online social networks and travel websites, in addition to climate data, can be used to support the inference of visitation count for many touristic attractions. To test our hypothesis we analyze visitation, climate and social media data in more than 70 National Parks in U.S during the last 3 years. The experimental results reveal a high correlation between social media data and tourism demands; in fact, in over 80\% of the parks, social media reviews and visitation counts are correlated by more than 50\%. Moreover, we assess the effectiveness of employing various prediction techniques, finding that even a simple linear regression model, when fed with social media and climate data as input features, can attain a prediction accuracy of over 80\% while a more robust algorithm, such as Support Vector Regression, reaches up to 94\% accuracy.


Improving Tourism Prediction Models Using Climate and Social Media Data: A Fine-Grained Approach

AAAI Conferences

Accurate predictions about future events is essential in many areas, one of them being the Tourism Industry. Usually, countries and cities invest a huge amount of money in planning and preparation in order to welcome (and profit from) tourists. An accurate prediction of the number of visits in the following days or months could help both the economy and tourists. Prior studies in this domain explore forecasting for a whole country rather than for fine-grained areas within a country (e.g., specific touristic attractions). In this work, we suggest that accessible data from online social networks and travel websites, in addition to climate data, can be used to support the inference of visitation count for many touristic attractions. To test our hypothesis we analyze visitation, climate and social media data in more than 70 National Parks in U.S during the last 3 years. The experimental results reveal a high correlation between social media data and tourism demands; in fact, in over 80\% of the parks, social media reviews and visitation counts are correlated by more than 50\%. Moreover, we assess the effectiveness of employing various prediction techniques, finding that even a simple linear regression model, when fed with social media and climate data as input features, can attain a prediction accuracy of over 80\% while a more robust algorithm, such as Support Vector Regression, reaches up to 94\% accuracy.


Li

AAAI Conferences

Although previous work has shown that member and structural features are important to the future popularity of groups in EBSN, it is not yet clear how different member roles and the interplay between them contribute to group popularity. In this paper, we study a real-world dataset from Meetup --- a popular EBSN platform --- and propose a deep neural network based method to predict the popularity of new Meetup groups. Our method uses group-level features specific to event-based social networks, such as time and location of events in a group, as well as the structural features internal to a group, such as the inferred member roles in a group and social substructures among members. Empirically, our approach reduces the RMSE of the popularity prediction (measured in RSVPs) of a group's future events by up to 12%, against the state-of-the-art baselines.


Akbari

AAAI Conferences

Online communities provide a unique way for individuals to access information from those in similar circumstances, which can be critical for health conditions that require daily and personalized management. As these groups and topics often arise organically, identifying the types of topics discussed is necessary to understand their needs. As well, these communities and people in them can be quite diverse, and existing community detection methods have not been extended towards evaluating these heterogeneities. This has been limited as community detection methodologies have not focused on community detection based on semantic relations between textual features of the user-generated content. Thus here we develop an approach, NeuroCom, that optimally finds dense groups of users as communities in a latent space inferred by neural representation of published contents of users. By embedding of words and messages, we show that NeuroCom demonstrates improved clustering and identifies more nuanced discussion topics in contrast to other common unsupervised learning approaches.