Showing 1–10 of 10 results for author: Howe, T

AI Chat Assistants can Improve Conversations about Divisive Topics

Authors: Lisa P. Argyle, Ethan Busby, Joshua Gubler, Chris Bail, Thomas Howe, Christopher Rytting, David Wingate

Abstract: A rapidly increasing amount of human conversation occurs online. But divisiveness and conflict can fester in text-based interactions on social media platforms, in messaging apps, and on other digital forums. Such toxicity increases polarization and, importantly, corrodes the capacity of diverse societies to develop efficient solutions to complex social problems that impact everyone. Scholars and c… ▽ More A rapidly increasing amount of human conversation occurs online. But divisiveness and conflict can fester in text-based interactions on social media platforms, in messaging apps, and on other digital forums. Such toxicity increases polarization and, importantly, corrodes the capacity of diverse societies to develop efficient solutions to complex social problems that impact everyone. Scholars and civil society groups promote interventions that can make interpersonal conversations less divisive or more productive in offline settings, but scaling these efforts to the amount of discourse that occurs online is extremely challenging. We present results of a large-scale experiment that demonstrates how online conversations about divisive topics can be improved with artificial intelligence tools. Specifically, we employ a large language model to make real-time, evidence-based recommendations intended to improve participants' perception of feeling understood in conversations. We find that these interventions improve the reported quality of the conversation, reduce political divisiveness, and improve the tone, without systematically changing the content of the conversation or moving people's policy attitudes. These findings have important implications for future research on social media, political deliberation, and the growing community of scholars interested in the place of artificial intelligence within computational social science. △ Less

Submitted 20 October, 2023; v1 submitted 14 February, 2023; originally announced February 2023.

Interpretable Classification of Bacterial Raman Spectra with Knockoff Wavelets

Authors: Charmaine Chia, Matteo Sesia, Chi-Sing Ho, Stefanie S. Jeffrey, Jennifer Dionne, Emmanuel J. Candès, Roger T. Howe

Abstract: Deep neural networks and other sophisticated machine learning models are widely applied to biomedical signal data because they can detect complex patterns and compute accurate predictions. However, the difficulty of interpreting such models is a limitation, especially for applications involving high-stakes decision, including the identification of bacterial infections. In this paper, we consider f… ▽ More Deep neural networks and other sophisticated machine learning models are widely applied to biomedical signal data because they can detect complex patterns and compute accurate predictions. However, the difficulty of interpreting such models is a limitation, especially for applications involving high-stakes decision, including the identification of bacterial infections. In this paper, we consider fast Raman spectroscopy data and demonstrate that a logistic regression model with carefully selected features achieves accuracy comparable to that of neural networks, while being much simpler and more transparent. Our analysis leverages wavelet features with intuitive chemical interpretations, and performs controlled variable selection with knockoffs to ensure the predictors are relevant and non-redundant. Although we focus on a particular data set, the proposed approach is broadly applicable to other types of signal data for which interpretability may be important. △ Less

Submitted 1 May, 2021; v1 submitted 8 June, 2020; originally announced June 2020.

Comments: 9 pages, 6 figures, 4 tables

Legal Area Classification: A Comparative Study of Text Classifiers on Singapore Supreme Court Judgments

Authors: Jerrold Soh Tsin Howe, Lim How Khang, Ian Ernst Chai

Abstract: This paper conducts a comparative study on the performance of various machine learning (``ML'') approaches for classifying judgments into legal areas. Using a novel dataset of 6,227 Singapore Supreme Court judgments, we investigate how state-of-the-art NLP methods compare against traditional statistical models when applied to a legal corpus that comprised few but lengthy documents. All approaches… ▽ More This paper conducts a comparative study on the performance of various machine learning (``ML'') approaches for classifying judgments into legal areas. Using a novel dataset of 6,227 Singapore Supreme Court judgments, we investigate how state-of-the-art NLP methods compare against traditional statistical models when applied to a legal corpus that comprised few but lengthy documents. All approaches tested, including topic model, word embedding, and language model-based classifiers, performed well with as little as a few hundred judgments. However, more work needs to be done to optimize state-of-the-art methods for the legal domain. △ Less

Submitted 12 April, 2019; originally announced April 2019.

Comments: Accepted to the 1st Workshop on Natural Legal Language Processing (co-located with NAACL2019)

Deep Learning for Lung Cancer Detection: Tackling the Kaggle Data Science Bowl 2017 Challenge

Authors: Kingsley Kuan, Mathieu Ravaut, Gaurav Manek, Huiling Chen, Jie Lin, Babar Nazir, Cen Chen, Tse Chiang Howe, Zeng Zeng, Vijay Chandrasekhar

Abstract: We present a deep learning framework for computer-aided lung cancer diagnosis. Our multi-stage framework detects nodules in 3D lung CAT scans, determines if each nodule is malignant, and finally assigns a cancer probability based on these results. We discuss the challenges and advantages of our framework. In the Kaggle Data Science Bowl 2017, our framework ranked 41st out of 1972 teams. We present a deep learning framework for computer-aided lung cancer diagnosis. Our multi-stage framework detects nodules in 3D lung CAT scans, determines if each nodule is malignant, and finally assigns a cancer probability based on these results. We discuss the challenges and advantages of our framework. In the Kaggle Data Science Bowl 2017, our framework ranked 41st out of 1972 teams. △ Less

Submitted 26 May, 2017; originally announced May 2017.

Active Control of Probability Amplitudes in a Mesoscale System via Feedback-Induced Suppression of Dissipation and Noise

Authors: Chaitanya Gupta, Aldo Peña Perez, Sean Fischer, Stephen Weinreich, Boris Murmann, Roger T. Howe

Abstract: We introduce a classical potentiostatic feedback mechanism that attenuates the dissipation in a quantum system arising from coupling to the surrounding thermodynamic bath, preserving the inter-state interference in an electronic excitation transfer (EET) process. A three-terminal potentiostat device applies a low-noise voltage bias to the terminals of the EET system and reduces the physical coupli… ▽ More We introduce a classical potentiostatic feedback mechanism that attenuates the dissipation in a quantum system arising from coupling to the surrounding thermodynamic bath, preserving the inter-state interference in an electronic excitation transfer (EET) process. A three-terminal potentiostat device applies a low-noise voltage bias to the terminals of the EET system and reduces the physical coupling between the quantum system and its environment. We introduce a classical equivalent circuit to model the environment-coupled excitation transfer in an elementary two-state system. This model provides qualitative insight into how classical feedback action affects the transition probabilities between the states and selectively reduces the dissipative coupling for one of the vibronic energy levels of the transfer system. Furthermore, we show that negative feedback results in persistent spectral coherence between the energy level of the decoupled state and the vibronic levels of the complementary state, making the decoupled vibronic channel a probe for characterizing the vibronic structure of the complementary channel of the EET system. △ Less

Submitted 26 May, 2016; originally announced May 2016.

Comments: 4 figures

Surfactant-aided exfoliation of molydenum disulphide for ultrafast pulse generation through edge-state saturable absorption

Authors: Richard C. T. Howe, Robert I. Woodward, Guohua Hu, Zongyin Yang, Edmund J. R. Kelleher, Tawfique Hasan

Abstract: We use liquid phase exfoliation to produce dispersions of molybdenum disulphide (MoS2) nanoflakes in aqueous surfactant solutions. The chemical structures of the bile salt surfactants play a crucial role in the exfoliation and stabilization of MoS2. The resultant MoS2 dispersions are heavily enriched in single and few (<6) layer flakes with large edge to surface area ratio. We use the dispersions… ▽ More We use liquid phase exfoliation to produce dispersions of molybdenum disulphide (MoS2) nanoflakes in aqueous surfactant solutions. The chemical structures of the bile salt surfactants play a crucial role in the exfoliation and stabilization of MoS2. The resultant MoS2 dispersions are heavily enriched in single and few (<6) layer flakes with large edge to surface area ratio. We use the dispersions to fabricate free-standing polymer composite wide-band saturable absorbers to develop mode-locked and Q- switched fibre lasers, tunable from 1535-1565 and 1030-1070 nm, respectively. We attribute this sub-bandgap optical absorption and its nonlinear saturation behaviour to edge-mediated states introduced within the material band-gap of the exfoliated MoS2 nanoflakes. △ Less

Submitted 11 December, 2015; v1 submitted 7 August, 2015; originally announced August 2015.

Comments: 6 pages, 5 figures

Yb- and Er-doped fiber laser Q-switched with an optically uniform, broadband WS2 saturable absorber

Authors: M. Zhang, G. Hu, G. Hu, R. C. T. Howe, L. Chen, Z. Zheng, T. Hasan

Abstract: We demonstrate a ytterbium (Yb) and an erbium (Er)-doped fiber laser Q-switched by a solution processed, optically uniform, few-layer tungsten disulfide saturable absorber (WS2-SA). Nonlinear optical absorption of the WS2-SA in the sub-bandgap region, attributed to the edge-induced states, is characterized by 3.1% and 4.9% modulation depths with 1.38 and 3.83 MW/cm2 saturation intensities at 1030… ▽ More We demonstrate a ytterbium (Yb) and an erbium (Er)-doped fiber laser Q-switched by a solution processed, optically uniform, few-layer tungsten disulfide saturable absorber (WS2-SA). Nonlinear optical absorption of the WS2-SA in the sub-bandgap region, attributed to the edge-induced states, is characterized by 3.1% and 4.9% modulation depths with 1.38 and 3.83 MW/cm2 saturation intensities at 1030 and 1558 nm, respectively. By integrating the optically uniform WS2-SA in the Yb- and Er-doped laser cavities, we obtain self-starting Q-switched pulses with microsecond duration and kilohertz repetition rates at 1030 and 1558 nm. Our work demonstrates broadband sub-bandgap saturable absorption of a single, solution processed WS2-SA, providing new potential efficacy for WS2 in ultrafast photonic applications. △ Less

Submitted 12 July, 2015; originally announced July 2015.

Wideband saturable absorption in few-layer molybdenum diselenide (MoSe2) for Q-switching Yb-, Er- and Tm-doped fiber lasers

Authors: R. I. Woodward, R. C. T. Howe, T. H. Runcorn, G. Hu, F. Torrisi, E. J. R. Kelleher, T. Hasan

Abstract: We fabricate a free-standing molybdenum diselenide (MoSe2) saturable absorber by embedding liquid-phase exfoliated few-layer MoSe2 flakes into a polymer film. The MoSe2-polymer composite is used to Q-switch fiber lasers based on ytterbium (Yb), erbium (Er) and thulium (Tm) gain fiber, producing trains of microsecond-duration pulses with kilohertz repetition rates at 1060 nm, 1566 nm and 1924 nm, r… ▽ More We fabricate a free-standing molybdenum diselenide (MoSe2) saturable absorber by embedding liquid-phase exfoliated few-layer MoSe2 flakes into a polymer film. The MoSe2-polymer composite is used to Q-switch fiber lasers based on ytterbium (Yb), erbium (Er) and thulium (Tm) gain fiber, producing trains of microsecond-duration pulses with kilohertz repetition rates at 1060 nm, 1566 nm and 1924 nm, respectively. Such operating wavelengths correspond to sub-bandgap saturable absorption in MoSe2, which is explained in the context of edge-states, building upon studies of other semiconducting transition metal dichalcogenide (TMD)-based saturable absorbers. Our work adds few-layer MoSe2 to the growing catalog of TMDs with remarkable optical properties, which offer new opportunities for photonic devices. △ Less

Submitted 12 June, 2015; v1 submitted 27 March, 2015; originally announced March 2015.

Journal ref: Opt. Express 23, 20051 (2015)

pMSSM Dark Matter Searches on Ice

Authors: R. C. Cotta, K. T. K. Howe, J. L. Hewett, T. G. Rizzo

Abstract: We explore the capability of the IceCube/Deepcore array to discover signal neutrinos resulting from the annihilations of Supersymmetric WIMPS that may be captured in the solar core. In this analysis, we use a previously generated set of /sim 70k model points in the 19-dimensional parameter space of the pMSSM which satisfy existing experimental and theoretical constraints. Our calculations employ a… ▽ More We explore the capability of the IceCube/Deepcore array to discover signal neutrinos resulting from the annihilations of Supersymmetric WIMPS that may be captured in the solar core. In this analysis, we use a previously generated set of /sim 70k model points in the 19-dimensional parameter space of the pMSSM which satisfy existing experimental and theoretical constraints. Our calculations employ a realistic estimate of the IceCube/DeepCore effective area that has been modeled by the IceCube collaboration. We find that a large fraction of the pMSSM models are shown to have significant signal rates in the anticipated IceCube/DeepCore 1825 day dataset, including some prospects for an early discovery. Many models where the LSP only constitutes a small fraction of the total dark matter relic density are found to have observable rates. We investigate in detail the dependence of the signal neutrino fluxes on the LSP mass, weak eigenstate composition, annihilation products and thermal relic density, as well as on the spin-independent and spin-dependent scattering cross sections. Lastly, We compare the model coverage of IceCube/DeepCore to that obtainable in near-future direct detection experiments and to pMSSM searches at the 7 TeV LHC. △ Less

Submitted 5 May, 2011; originally announced May 2011.

Comments: 37 pages, 17 figures

Report number: SLAC-PUB-14390, SU-ITP-11/06

Tensile-Strained Germanium-on-Insulator Substrate Fabrication for Silicon-Compatible Optoelectronics

Authors: J. Raja Jain, Dany-Sebastien Ly-Gagnon, Krishna C. Balram, Justin S. White, Mark L. Brongersma, David A. B. Miller, Roger T. Howe

Abstract: We present a method to fabricate tensile-strained germanium-on-insulator (GOI) substrates using heteroepitaxy and layer transfer techniques. The motivation is to obtain a high-quality wafer-scale GOI platform suitable for silicon-compatible optoelectronic device fabrication. Crystal quality is assessed using X-Ray Diffraction (XRD) and Transmission Electron Microscopy. A biaxial tensile film strai… ▽ More We present a method to fabricate tensile-strained germanium-on-insulator (GOI) substrates using heteroepitaxy and layer transfer techniques. The motivation is to obtain a high-quality wafer-scale GOI platform suitable for silicon-compatible optoelectronic device fabrication. Crystal quality is assessed using X-Ray Diffraction (XRD) and Transmission Electron Microscopy. A biaxial tensile film strain of 0.16% is verified by XRD. Suitability for device manufacturing is demonstrated through fabrication and characterization of metal-semiconductor-metal photodetectors that exhibit photoresponse beyond 1.55 μm. The substrate fabrication process is compatible with complementary metal-oxide-semiconductor manufacturing and represents a potential route to wafer-scale integration of silicon-compatible optoelectronics. △ Less

Submitted 17 September, 2011; v1 submitted 30 April, 2011; originally announced May 2011.

Journal ref: Optical Materials Express 1(6), 1121--1126 (2011)