Organizational Unit:

School of Interactive Computing

Permanent Link

https://hdl.handle.net/1853/70783

Parent Organization

Organizational Unit

College of Computing

ArchiveSpace Name Record

https://finding-aids.library.gatech.edu/agents/corporate_entities/1113

Full item page

Publication Search Results

Now showing 1 - 10 of 295

Robot Manipulation Alongside and in Collaboration with People

(Georgia Institute of Technology, 2021-01-22) Kent, David ; Chernova, Sonia ; Gombolay, Matthew ; Kemp, Charlie ; Shah, Julie ; Fong, Terry ; Interactive Computing

Autonomous robot manipulation in unstructured environments is a required behavior for many robotics applications, from day-to-day household tasks to remote exploration of dangerous environments. To effectively deploy such systems to solve real-world problems requires approaches and representations that are compatible with people, who can be direct collaborators, subjects of assistance, or independent agents sharing the robot’s environment. The objective of this work is to improve the autonomy of robot manipulators in unstructured environments while addressing the unique challenges of working with and around humans. To address such challenges, we posit that autonomous manipulators must be both easily adjustable by system designers, and adaptive to humans in the environment, which we achieve through the use of transparent representations, human-in-the-loop systems, and learning from demonstration, across both skill- and task-level manipulation. This thesis seeks to investigate the hypothesis that improved autonomy, adjustability of behavior, and adaptiveness to people lead to greater robot efficiency and effectiveness in manipulation tasks when operating alongside and in collaboration with people. To support this claim, this thesis contributes: (1) novel approaches for human-in-the-loop grasp pose specification for teleoperation that leverage depth data and robot autonomy to balance responsibilities between the operator and the robot; (2) efficient skill-level learning by means of a pairwise ranking formulation of autonomous grasp calculation that enables robust mobile manipulation and supports interaction-efficient training and adaptability; (3) efficient task-level learning by means of a novel unsupervised learning approach for hierarchical task models with action execution preferences that enable human-robot collaboration; (4) a novel algorithm for adaptive and collaborative task planning that builds on our learned hierarchical task representation; and (5) a formulation and exploration of autonomous human observation that utilizes manipulation-enabled free-flying robots to unobtrusively support humans during non-collaborative tasks in remote environments.
THE PRIVATE SECTOR AND CIVIL CONFLICT: LEVERAGING ECONOMIC SECTORS FOR PEACE

(Georgia Institute of Technology, 2021-12-13) Shabb, Rana Olivia ; Rubin, Lawrence ; Kosal, Margaret ; Thomas, Anjali ; Woodall, Brian ; Diwan, Ishac ; International Affairs

My dissertation, The Private Sector and Civil Conflict: Leveraging Economic Sectors for Peace, aims to better understand the relationship between the private sector and peace, filling an academic gap and addressing a policy need. To this end, I undertake three interrelated research projects. The inquiries put forth and their findings seek to leverage existing policy tools to help strengthen peacebuilding and conflict prevention interventions in countries affected by conflict. To frame the questions undertaken in this dissertation, I provide a quick overview in Chapter 1 covering the main bodies of literature that seek to understand peace and prosperity in developing countries. While metatheories of modernization, liberalism, and intuitionalism have sought grand-scale explanations, they tend to assume that capitalism is part and parcel of peaceful and prosperous societies. The private sector is treated as a black box. At a more granular level, this dissertation seeks to understand how capitalism and the private sector affect the civil conflict-peace dynamic. Further, more targeted civil conflict academic work – which is more positivist and exhaustive in nature – tends to highlight economic factors (economic growth, poverty, price shocks) as drivers for conflict. Nevertheless, there is significantly less examination or theorization of how the private sector and firms can contribute to these factors. Previous approaches treat the private sector as a consumer of its environment (in terms of property rights, labor, prices, privileges). As such, this dissertation fills an analytical and theoretical gap and shifts the level of analysis to the private sector and firm-level. From this perspective, the private sector engages with labor (would-be-rebels and those with possible grievances) and governments to advance their material interest. Better understanding the private sector- civil conflict nexus sheds light on previously unexamined areas and can help inform peacebuilding interventions in developing countries. Notwithstanding academic work, conventional wisdom in the practitioner community states that a vibrant private sector is necessary to help secure peace in conflict-affected countries. International development agencies, for instance, have adopted private sector development as a strategy to promote peace. Despite this conviction, there is little to no evidence in the academic literature to support this claim. In chapter 2, I draw on the business and peace, and civil conflict literatures, to argue that a strong private sector through job creation and growth decreases prospective rebels’ incentives to join a rebellion and eventually reduces the likelihood of civil conflict. The argument is tested by examining the effects of private sector strength, as measured by domestic credit granted to the private sector and investment climate, on the probability of civil conflict occurrence from 1995 through 2018. Statistical analysis shows that a strong private sector has a pacifying effect on civil conflict. Specifically, findings demonstrate that access to credit, rather than investment climate, is more effective at sustaining peace. I illustrate the quantitative findings with the comparative cases of Egypt and Tunisia to show the mechanism by which access to credit has higher peace dividends. These findings fill an academic gap and equip policymakers to make more effective peacebuilding interventions. Further, the civil conflict literature tends to compare conflict nationally and does not differentiate between economic sectors, with the exception of the extractive industries. In chapter 3, I address the question of whether some economic sectors are better than others at sustaining peace. To examine firms’ subnational contributions to peace or civil conflict, I build a theoretical framework to predict economic sectors’ propensity for peace. Based on the supply of factors of production in civil conflict, I deduce that economic sectors that rely on skilled labor, mobile, and high-tech equipment are more vulnerable than those that rely on unskilled, fixed, and low-tech equipment. Subsequently, I argue that firms operating in sectors vested in peace (for their bottom line) engage in peace-promoting activities. To test for differentiated effects, I conduct a focused and structured within case analysis in Lebanon examining two sectors: one vested in peace and the other peace-neutral (financial vs. quarrying sector). Analysis of fieldwork data, collected through semi-structured interviews and local news reports reveals that firms vested in peace support national policies to that effect, whereas peace-neutral business can engage in inflammatory tactics, which have occasionally led to violent conflict. Given that knowledge and high-tech intensive economic sectors are more vested in peace than others, can existing foreign policy tools be leveraged to promote innovation in recipient economies? Chapter 4 examines the conditions under which military aid to developing countries triggers innovation. This question emanates from a puzzle in the innovation literature. Studies focused on military expenditure in the developed world show a positive relationship between military expenditure and innovation. Conversely, studies centered on military expenditure in developing countries often note the unintended, negative consequences of such expenditure (autocracy, increased coups, and the undermining of human rights). Borrowing from current literature on innovation that examines diffusion channels from the military to the national economy, this research seeks to identify a similar process in developing countries. Using a congruence test on a least-likely case, this study finds that military aid – effectively a military expenditure subsidy – can indeed trigger the emergence of new high-tech knowledge intensive sector in a recipient economy. In Jordan, this is reflected as the emergence of an innovative domestic arms industry after its peace agreement with Israel and a major influx of U.S. military aid. Further, by dividing military aid into different sub-types and tracing and comparing their different effects, this study finds that with conducive industrial and S&T domestic policy, military aid can have secondhand virtuous effects and lead to innovation in the recipient economy. Finally, Chapter 5 concludes by highlighting the main findings from the project, policy recommendations, and avenues for future research. Overall, this dissertation sheds light on how the private sector can help sustain peace, and how military aid – already dispatched in the billions – can be leveraged to magnify virtuous second-hand effects that work to support peace and prosperity in the long run.
User’s role in shaping WeChat as an infrastructure: practice, appropriation, creation

(Georgia Institute of Technology, 2020-08-17) Zhou, Rui RZ ; DiSalvo, Betsy ; Grinter, Rebecca E. ; DiSalvo, Carl ; Anderson, Ken ; Gui, Xinning ; Interactive Computing

The past decade has seen the rapid development of information and communication technologies, particularly online social platforms such as Twitter and Facebook. While online social platforms have existed since the early years of the internet, it is only in recent years that they begin to set foot on mobile devices, bringing them accessible to more people from all over the globe. Gradually gaining their presence in people’s everyday lives, some of these social platforms have started expanding themselves from a simple social platform to a more powerful, more embedded, and more transparent infrastructure, supporting their users through various ways that are not limited to social or communication aspects. One instance of such a social platform that has successfully turned into an infrastructure is WeChat, the most popular mobile social application in China. Introduced in 2011, WeChat is currently the fifth largest social networking platform in the world, holding 1.2 billion monthly active users. When it was first developed, WeChat was solely a mobile instant messenger that supported users with a set of common communication media. However, through its growth in the past nine years, it has also designed and integrated many non-communication, non-social functions for online payment, gaming, and much more. Nowadays, Chinese people use WeChat all the time: from paying street vendors to calling ride-hailing services, from reading daily news to reserving restaurant tables, WeChat is not only a communication tool but also an all-encompassing platform and infrastructure that Chinese people use to fulfill all kinds of needs. Given this prominent presence of WeChat and its status as both a platform and an infrastructure, WeChat’s development and its relationship with its users are worth studying; its success offers a lot to learn about other similar platforms that are swiftly evolving into infrastructures. This dissertation delves deep into understanding WeChat by focusing on how people use it. It asks questions about how people use WeChat, why people use WeChat, and how people’s use of WeChat has influenced WeChat to move from a platform to an infrastructure. To answer these questions, five empirical studies were conducted, revolving around Chinese people’s use of different functions on WeChat under various situations and scenarios. Relying on qualitative methods, these studies together provide a holistic view of how people use WeChat. In addition, a meta-analysis was done on data collected from these studies, aiming for teasing out the user’s role in WeChat’s evolvement from a platform to an infrastructure. Findings from these studies reveal that while WeChat influences users and shapes their interactions with each other, it is affected and changed by users’ practices as well. Furthermore, by using WeChat, users, knowingly or not, have pushed WeChat to become a powerful infrastructure. This dissertation is the first in-depth study that researches diverse aspects of WeChat by attending to people’s ways of using it, providing a holistic view of Chinese people’s engagements with WeChat in their everyday lives and how these engagements contribute to WeChat’s infrastructuralization. This dissertation offers three major contributions to the field of Human-Computer Interaction: first, it provides an in-depth exploration of a popular non-Western social and communication application; second, by taking a user-centered perspective, this dissertation uncovers the mutual-shaping relationship between WeChat and its users; third, most importantly, this dissertation contributes to understanding other social platforms’ infrastructuralization processes by using WeChat as an exemplar, uncovering the role played by users in this process.
Creating Necessary Knowledge

(Georgia Institute of Technology, 2004) Wilson, Ernest J., III ; Best, Michael L. ; Massachusetts Institute of Technology ; University of Southern California. Annenberg School for Communication & Journalism ; Georgia Institute of Technology. School of International Affairs ; Georgia Institute of Technology. School of Interactive Computing
Managing learning interactions for collaborative robot learning

(Georgia Institute of Technology, 2019-09-11) Bullard, Kalesha ; Chernova, Sonia ; Isbell, Charles ; Christensen, Henrik I. ; Mataric, Maja ; Thomaz, Andrea L. ; Interactive Computing

Robotic assistants should be able to actively engage their human partner(s) to generalize knowledge about relevant tasks within their shared environment. Yet a key challenge is not all human partners will be proficient at teaching; furthermore, humans should not be held accountable for tracking a robot’s knowledge over time in a dynamically changing environment, across multiple tasks. Thus, it is important to enable these interactive robots to characterize their own uncertainty and equip them with an information gathering policy for asking the appropriate questions of their human partners to resolve that uncertainty. In this way, the robot shares the responsibility in guiding its own learning process and is a collaborator in the learning. Additionally, given the robot requires some tutelage from its partner, awareness of constraints on the teacher’s time and cognitive resources available for devoting to the interaction could help the agent to use the time allotted more wisely. This thesis examines the problem of enabling a robotic agent to leverage structured interaction with a human partner for acquiring concepts relevant to a task it must later perform. To equip the agent with the desired concept knowledge, we first explore the paradigm of Learning from Demonstration for the acquisition of (1) training instances as examples of task-relevant concepts and (2) informative features for appropriately representing and discriminating between task-relevant concepts. Given empirical evidence that a human partner can be helpful to the agent in solving the concept learning problem, we subsequently investigate the design of algorithms that enable the robot learner to autonomously manage interaction with its human partner, using a questioning policy to actively gather both instance and feature information. This thesis seeks to investigate the following hypothesis: In the context of robot learning from human demonstrations in changeable and resource-constrained environments, enabling the robot to actively elicit multiple types of information through questions, and to reason about what question to ask and when, leads to improved learning performance.
Supporting and transforming leadership in online creative collaboration

(Georgia Institute of Technology, 2012-08-24) Luther, Kurt ; Bruckman, Amy S. ; Ahamad, Mustaque ; Copeland, John ; Feamster, Nicholas G. ; Traynor, Patrick ; Computing

Online creative collaboration is challenging our basic assumptions about how people can create together. Volunteers from around the world who meet and communicate over the Internet have written the world's largest encyclopedia, developed market-leading software products, solved important open problems in mathematics, and produced award-winning films, among many examples. A growing body of research refutes the popular myth that these projects succeed through "self-organization" and instead points to the critical importance of effective leadership. Yet, we know little about what these leaders actually do, the challenges they must manage, and how technology supports or hinders their efforts. In this dissertation, I investigated the role of leadership in online creative collaboration. I first conducted two empirical studies of existing leadership practices, focusing on the domain of online, collaborative animation projects called "collabs." In the first study, I identified the major challenges faced by collab leaders. In the second study, I identified leader traits and behaviors correlated with success. These initial findings suggested that many collab leaders, overburdened and lacking adequate technological support, respond by attempting less ambitious projects and adopting centralized leadership styles. Despite these efforts, leaders frequently become overburdened, and more than 80% of collabs fail. To ease the burden on leaders and encourage more complex, successful projects, I led the development of a web-based, open-source software tool called Pipeline. Pipeline can support leadership by reinforcing a traditional, top-down approach, or transform leadership by redistributing it across many members of a group. This latter approach relies on social processes, rather than technical constraints, to guide behavior. I evaluated Pipeline's ability to effectively support and transform leadership through a detailed case study of Holiday Flood, a six-week collaboration involving nearly 30 artists from around the world. The case study showed that formal leaders remained influential and Pipeline supported their traditional, centralized approach. However, there was also evidence that Pipeline transformed some leadership behaviors, such as clarifying, informing, and monitoring, by redistributing them beyond the project's formal leaders. The result was a significantly more ambitious project which attained its goals and earned high praise from the community. The main contributions of this dissertation include: (1) a rich description of existing leadership practices in online creative collaboration; (2) the development of redistributed leadership as a theoretical framework for analyzing the relationship between leadership and technological support; (3) design implications for supporting and transform leadership; (4) a case study illustrating how technology can support and transform leadership in the real world; and (5) the Pipeline collaboration tool itself, released as open-source software.
Disentangling neural network representations for improved generalization

(Georgia Institute of Technology, 2020-04-24) Cogswell, Michael Andrew ; Batra, Dhruv ; Parikh, Devi ; Hays, James ; Goel, Ashok ; Lee, Stefan ; Interactive Computing

Despite the increasingly broad perceptual capabilities of neural networks, applying them to new tasks requires significant engineering effort in data collection and model design. Generally, inductive biases can make this process easier by leveraging knowledge about the world to guide neural network design. One such inductive bias is disentanglment, which can help preven neural networks from learning representations that capture spurious patterns that do not generalize past the training data, and instead encourage them to capture factors of variation that explain the data generally. In this thesis we identify three kinds of disentanglement, implement a strategy for enforcing disentanglement in each case, and show that more general representations result. These perspectives treat disentanglement as statistical independence of features in image classification, language compositionality in goal driven dialog, and latent intention priors in visual dialog. By increasing the generality of neural networks through disentanglement we hope to reduce the effort required to apply neural networks to new tasks and highlight the role of inductive biases like disentanglement in neural network design.
Towards a Canine-Human Communication System Based on Head Gestures

(Georgia Institute of Technology, 2015-11) Valentin, Giancarlo ; Alcaidinho, Joelle ; Howard, Ayanna M. ; Jackson, Melody Moore ; Starner, Thad ; Georgia Institute of Technology. College of Computing ; Georgia Institute of Technology. School of Interactive Computing ; Georgia Institute of Technology. School of Electrical and Computer Engineering

We explored symbolic canine-human communication for working dogs through the use of canine head gestures. We identified a set of seven criteria for selecting head gestures and identified the first four deserving further experimentation. We devised computationally inexpensive mechanisms to prototype the live system from a motion sensor on the dog’s collar. Each detected gesture is paired with a predetermined message that is voiced to the humans by a smart phone. We examined the system and proposed gestures in two experiments, one indoors and one outdoors. Experiment A examined both gesture detection accuracy and a dog’s ability to perform the gestures using a predetermined routine of cues. Experiment B examined the accuracy of this system on two outdoor working-dog scenarios. The detection mechanism we presented is sufficient to point to improvements into system design and provide valuable insights into which gestures fulfill the seven minimum criteria.
The role of copyright in online creative communities: law, norms, and policy

(Georgia Institute of Technology, 2015-07-23) Fiesler, Casey ; Bruckman, Amy S. ; Anton, Annie ; Gilbert, Eric ; Lampe, Cliff ; Tushnet, Rebecca ; Interactive Computing

Many sources of rules govern our interactions with technology and our behavior online—law, ethical guidelines, community norms, website policies—and they do not always agree. This is particularly true in the context of content production because copyright law represents a collection of complex policies that often do not always account for the ways that people use and re-use digital media. Within legal gray areas, people make decisions every day about what is allowed, often negotiating multiple sources of rules. How do content creators make decisions about what they can and cannot do when faced with unclear rules, and how does the law (and perceptions of the law) impact technology use, creativity, and online interaction? Combining in-depth interviews, large-scale content analysis, and surveys, my work examines the complex relationship between law, site policy, norms, and technology. This dissertation provides a better understanding of how content creators engage with copyright and how norms organically form within communities of creators. It concludes with a set of design and policy recommendations for online community designers to help better support current practices among content creators.
A data-driven approach for personalized drama management

(Georgia Institute of Technology, 2015-05-13) Yu, Hong ; Riedl, Mark O. ; Isbell, Charles ; Magerko, Brian ; Roberts, David ; Thomaz, Andrea ; Interactive Computing

An interactive narrative is a form of digital entertainment in which players can create or influence a dramatic storyline through actions, typically by assuming the role of a character in a fictional virtual world. The interactive narrative systems usually employ a drama manager (DM), an omniscient background agent that monitors the fictional world and determines what will happen next in the players' story experience. Prevailing approaches to drama management choose successive story plot points based on a set of criteria given by the game designers. In other words, the DM is a surrogate for the game designers. In this dissertation, I create a data-driven personalized drama manager that takes into consideration players' preferences. The personalized drama manager is capable of (1) modeling the players' preference over successive plot points from the players' feedback; (2) guiding the players towards selected plot points without sacrificing players' agency; (3) choosing target successive plot points that simultaneously increase the player's story preference ratings and the probability of the players selecting the plot points. To address the first problem, I develop a collaborative filtering algorithm that takes into account the specific sequence (or history) of experienced plot points when modeling players' preferences for future plot points. Unlike the traditional collaborative filtering algorithms that make one-shot recommendations of complete story artifacts (e.g., books, movies), the collaborative filtering algorithm I develop is a sequential recommendation algorithm that makes every successive recommendation based on all previous recommendations. To address the second problem, I create a multi-option branching story graph that allows multiple options to point to each plot point. The personalized DM working in the multi-option branching story graph can influence the players to make choices that coincide with the trajectories selected by the DM, while gives the players the full agency to make any selection that leads to any plot point in their own judgement. To address the third problem, the personalized DM models the probability that the players transitioning to each full-length stories and selects target stories that achieve the highest expected preference ratings at every branching point in the story space. The personalized DM is implemented in an interactive narrative system built with choose-your-own-adventure stories. Human study results show that the personalized DM can achieve significantly higher preference ratings than non-personalized DMs or DMs with pre-defined player types, while preserve the players' sense of agency.