• Journal of Information Processing Systems
• /
• v.19 no.5
• /
• pp.641-651
• /
• 2023

Malicious hate speech and gender bias comments are common in online communities, causing social problems in our society. Gender bias and hate speech detection has been investigated. However, it is difficult because there are diverse ways to express them in words. To solve this problem, we attempted to detect malicious comments in a Korean hate speech dataset constructed in 2020. We explored bidirectional encoder representations from transformers (BERT)-based deep learning models utilizing hyperparameter tuning, data sampling, and logits ensembles with a label distribution. We evaluated our model in Kaggle competitions for gender bias, general bias, and hate speech detection. For gender bias detection, an F1-score of 0.7711 was achieved using an ensemble of the Soongsil-BERT and KcELECTRA models. The general bias task included the gender bias task, and the ensemble model achieved the best F1-score of 0.7166.

• Progress in Superconductivity
• /
• v.14 no.2
• /
• pp.96-98
• /
• 2012

Bias current dependence of a superconducting strip photon detector is studied in the wavelength range of 405 to 1310 nm. The detector is made of an $MgB_2$ meander pattern with the line width of 135 nm and thickness of 10 nm. At 1310 nm, the detection efficiency exponentially decreases as the bias current decreases. While at 405 nm, the detection efficiency almost saturates in the high bias current region. These features suggest that the intrinsic detection efficiency of the $MgB_2$ detector is high at 405 nm.

• 제어로봇시스템학회:학술대회논문집
• /
• 1992.10a
• /
• pp.554-557
• /
• 1992

In this paper the normalized innovations squared of a Kalman filter is used to detect a bias level in prediction errors due to target accelerations. The probability density function of the normalized innovation squared is obtained for a steady state Kalman filter, and it is used to calculate the detection probability of the bias level. A typical example is given to compute the detection probability.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.36 no.9
• /
• pp.678-684
• /
• 1987

The identification problem of instrument faults in PWR pressurizer is considered. The instrument failure detection technique in this paper consists of two filters, a normal-mode Kalman filter which estimates plant states in normal operation and a bias estimator which estimates the magnitudes and directions of bias faults. The concept of threshold based on the residual of a Kalman filter in normal operation is introduced. The bias estimator is driven when the absolute value of residual exceeds the threshold. The suggested failure detection algorithm is applied to a PWR pressurizer. Computer simulations show that the prompt detection of bias fault can be performed very successfully when there exist instrument faults.

• Proceedings of the Korea Electromagnetic Engineering Society Conference
• /
• 2005.11a
• /
• pp.97-100
• /
• 2005

This paper proposes a linearization method that is to control the operating point of a class AB amplifier according to its output power. The proposed linearization method is presented in this paper and the performance test results using two-tone signal are presented also.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.86.5-86
• /
• 2002

The measurement sensor may contain unknown bias in addition to the white noise in the measurement sequence. In this paper, fault detection and diagnosis scheme for the measurement sensor is developed based on the adaptive estimator. The proposed scheme consists of a parallel bank of Kalman-type filters each matched to a set of different possible biases, a mode probability evaluator, an estimate combiner at the outputs of the filters, a bias estimator, and a fault detection and diagnosis logic. Monte Carlo simulations for the PWR steam generator in the nuclear power plant are provided to illustrate the effectiveness of the proposed scheme.

• Journal of Information Processing Systems
• /
• v.9 no.4
• /
• pp.538-547
• /
• 2013

In this paper, we propose a bias detection method that is based on personal and social opinions that express contrasting views on competing topics on Twitter. We used unsupervised polarity classification is conducted for learning social opinions on targets. The $tf{\cdot}idf$ algorithm is applied to extract targets to reflect sentiments and features of tweets. Our method addresses there being a lack of a sentiment lexicon when learning social opinions. To evaluate the effectiveness of our method, experiments were conducted on four issues using Twitter test collection. The proposed method achieved significant improvements over the baselines.

• Journal of Sensor Science and Technology
• /
• v.2 no.1
• /
• pp.57-64
• /
• 1993

In this paper the normalized innovations squared of a Kalman filter is used to detect a bias level in prediction errors due to target accelerations. The probability density function of the normalized innovation squared is obtained for a steady state Kalman filter, and it is used to calculate the detection probability of the bias level. A typical example is given to compute the detection probability and to plot the maneuver detector operating characteristic curves.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.32 no.9
• /
• pp.88-93
• /
• 2004

A method is suggested for validating the existence of bias error on GPS carrier measurement. The baseline constraint is used as an addition measurement, which augments the original measurement equation. The detection probability is calculated on both cases. The first case is using GPS carrier measurement only, the second case is using GPS carrier and a baseline constraint. The improvement of the detection probability is shown, and the advantage of using baseline constraint is described statistically, the results of the simulation is shown and analyzed.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.4
• /
• pp.254-260
• /
• 2000

This paper deals with the development of fault tolerant control for a nonlinear boiler system with noise and disturbance. The MCMBPC(Multivariable Constrained Model Based Predictive Control) is adopted for the control of the specific boiler turbin model. The fault detection and diagnosis are accomplished with the Kalman filter and two bias estimators. Once a fault is detected, two Bias estimators are driven to estimate the fault and to discriminate Process fault and sensor fault. In this paper, a fault tolerant control scheme combining MCMBPC with a fault compensation method based on the bias estimator is proposed. The proposed scheme has been applied to the nonlinear boiler system and shown a satisfactory performance through some simulations.