PBPaul Bucci

Paul Bucci

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Paul Bucci has a total of 16 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Crystallizing Schemas with Teleoscope: Thematic Curation of Large Text Corpora on Reddit

    by Patrick Yung Kang Lee, Paul Hendrik Bucci, Leo Itsuki Foord-Kelcey, Alamjeet Singh, Ivan Beschastnikh

  • Open Access

    Making Sense of Multi-threaded Application Performance at Scale with NonSequitur

    by Augustine Wong, Paul Bucci, Ivan Beschastnikh, Alexandra Fedorova

    Modern multi-threaded systems are highly complex. This makes their behavior difficult to understand. Developers frequently capture behavior in the form of program traces and then manually inspect these traces. Existing tools, however, fail to scale to traces larger than a million events. In this paper we present an approach to compress multi-threaded traces in order to allow developers to visually explore these traces at scale. Our approach is able to compress traces that contain millions of events down to a few hundred events. We use this approach to design and implement a tool called NonSequitur. We present three case studies which demonstrate how we used NonSequitur to analyze real-world performance issues with Meta’s storage engine RocksDB and MongoDB’s storage engine WiredTiger, two complex database backends. We also evaluate NonSequitur with 42 participants on traces from RocksDB and WiredTiger. We demonstrate that, in some cases, participants on average scored 11 times higher when performing performance analysis tasks on large execution traces. Additionally, for some performance analysis tasks, the participants spent on average three times longer with other tools than with NonSequitur.


  • FEELing (key)Pressed: Implicit Touch Pressure Bests Brain Activity for Modeling Emotion Dynamics in the Space Between Stressed & Relaxed

    by X. Laura Cang, Rubia R. Guerra, Bereket Guta, Paul Bucci, Laura Rodgers, Hailey Mah, Qianqian Feng, Anushka Agrawal, Karon E. MacLean

    In-body lived emotional experiences can be complex, with time-varying and dissonant emotions evolving simultaneously; devices responding in real-time to estimate personal human emotion should evolve accordingly. Models assuming generalized emotions exist as discrete states fail to operationalize valuable information inherent in the dynamic and individualistic nature of human emotions. Our multi-resolution emotion self-reporting procedure allows the construction of emotion labels along the Stressed-Relaxed scale, differentiating not only what the emotions are, but how they are transitioning - e.g., "hopeful but getting stressed" vs. "hopeful and starting to relax". We trained participant-dependent hierarchical models of contextualized individual experience to compare emotion classification by modality (brain activity and keypress force from a physical keyboard), then benchmarked classification performance at F1-scores = [0.44, 0.82] (chance F1=0.22, sigma = 0.01) and examined high-performing features. Notably, when classifying emotion evolution in the context of an experience that realistically varies in stress, pressure-based features from keypress force proved to be the more informative modality, and more convenient when considering intrusiveness and ease of collection and processing. Finally, we present our FEEL (Force, EEG and Emotion-Labelled) dataset, a collection of brain activity and keypress force data, labelled with self-reported emotion collected during tense videogame play (N = 16) and open-sourced for community exploration.


  • Open Access

    Affective Robots Need Therapy

    by Paul Bucci, David Marino, Ivan Beschastnikh

    Emotion researchers have begun to converge on the theory that emotions are psychologically and socially constructed. A common assumption in affective robotics is that emotions are categorical brain-body states that can be confidently modeled. But if emotions are constructed, then they are interpretive, ambiguous, and specific to an individual’s unique experience. Constructivist views of emotion pose several challenges to affective robotics: first, it calls into question the validity of attempting to obtain objective measures of emotion through rating scales or biometrics. Second, ambiguous subjective data poses a challenge to computational systems that need structured and definite data to operate. How can a constructivist view of emotion be rectified with these challenges? In this article, we look to psychotherapy for ontological, epistemic, and methodological guidance. These fields (1) already understand emotions to be intrinsically embodied, relative, and metaphorical and (2) have built up substantial knowledge informed by everyday practice. It is our hope that by using interpretive methods inspired by therapeutic approaches, HRI researchers will be able to focus on the practicalities of designing effective embodied emotional interactions.


  • Discerning Affect From Touch and Gaze During Interaction With a Robot Pet

    by Xi Laura Cang, Paul Bucci, Jussi Rantala, Karon E. MacLean

    Practical affect recognition needs to be efficient and unobtrusive in interactive contexts. One approach to a robust real-time system is to sense and automatically integrate multiple nonverbal sources. We investigated how users' touch, and secondarily gaze, perform as affect-encoding modalities during physical interaction with a robot pet, in comparison to more-studied biometric channels. To elicit authentically experienced emotions, participants recounted two intense memories of opposing polarity in Stressed-Relaxed or Depressed-Excited conditions. We collected data (N=30) from a touch sensor embedded under robot fur (force magnitude and location), a robot-adjacent gaze tracker (location), and biometric sensors (skin conductance, blood volume pulse, respiration rate). Cross-validation of Random Forest classifiers achieved best-case accuracy for combined touch-with-gaze approaching that of biometric results: where training and test sets include adjacent temporal windows, subject-dependent prediction was 94% accurate. In contrast, subject-independent Leave-One-participant-Out predictions resulted in 30% accuracy (chance 25%). Performance was best where participant information was available in both training and test sets. Addressing computational robustness for dynamic, adaptive real-time interactions, we analyzed subsets of our multimodal feature set, varying sample rates and window sizes. We summarize design directions based on these parameters for this touch-based, affective, and hard, real-time robot interaction application.