• Journal of Computing Science and Engineering
• /
• v.3 no.3
• /
• pp.143-164
• /
• 2009

This article surveys recent research in the area of language modeling (sometimes called statistical language modeling) approaches to information retrieval. Language modeling is a formal probabilistic retrieval framework with roots in speech recognition and natural language processing. The underlying assumption of language modeling is that human language generation is a random process; the goal is to model that process via a generative statistical model. In this article, we discuss current research in the application of language modeling to information retrieval, the role of semantics in the language modeling framework, cluster-based language models, use of language modeling for XML retrieval and future trends.

• Journal of the Korean Society for information Management
• /
• v.10 no.2
• /
• pp.43-67
• /
• 1993

The primary purpose of the study is to improve the low preclslon often found In full-text retrleval systems. In order to enhance the low precision of full-text retrleval wh~le retaining ~ t s hgh recall, relevance feedback mechanisms based on probabilistic retrieval models (binary independence and two-Polsson Independence models) were employed. Thls paper investigates the effect of relevance feedback on the performance of full-text retrieval systems.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2005.05b
• /
• pp.249-252
• /
• 2005

Chloride ingress is a common cause of deterioration of reinforced concrete structures. Modeling the chloride ingress is an important basis for designing reinforced concrete structures and for assessing the reliability of an existing structure. The modelling is also needed for predicting the deterioration of a reinforced structure. This paper presents an approach for the probabilistic modeling of the chloride-induced corrosion of reinforcement steel in concrete structures that takes into account the uncertainties in the physical models. The parameters of the models are modeled as random variables and the distribution of the corrosion time and probability of corrosion are determined by using Monte Carlo simulation. The predictions of the proposed model is very effective to do the decision-making about initiation time and deterioration degree.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1997.04a
• /
• pp.27-34
• /
• 1997

In this paper, systematic and comprehensive approaches are suggested for the application of quantitative PRA techniques especially for those risk events that cannot be easily evaluated quantitatively In addition, dominant risk events are identified based on their occurrence frequency assessed by both actual survey of construction site conditions and the statistical data related with the probable accidents. Practical FTA(Fault Tree Analysis) and ETA(Event Tree Analysis) models are used for the assessment of the identified risks. When the risk events are lack of statistical data, appropriate Bayesian models incorporating engineering judgement and test results are also introduced in this paper. Moreover, a fuzzy probability technique is used for the quantitative risk assessment of those risk components which are difficult to evaluate quantitatively.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 2003.09a
• /
• pp.36-45
• /
• 2003

Sensitivity analyses for several seismic source models were studied. For the area sources, the hazard is steeply decreasing with the source-to-site distance. Hazard is decreasing when the area of the source is increasing with fixed annual rate. For the fault sources, the fault length, distance from a site and dip angle of near fault show very sensitive effect to seismic hazard. But the various magnitude-rupture length relationships show effect to seismic hazard slightly. For the fault source with small magnitude, the exponential model is preferred rather than the characteristic model to the magnitude-recurrence law.

• Nuclear Engineering and Technology
• /
• v.53 no.2
• /
• pp.376-385
• /
• 2021

Probabilistic risk assessment (PRA) has improved its elemental technologies used for assessing external events since the Fukushima Daiichi Nuclear Power Station Accident in 2011. HRA needs to be improved for analyzing tasks performed under extreme conditions (e.g., different actors responding to external events or performing operations using portable mitigation equipment). To make these improvements, it is essential to understand plant-specific and scenario-specific conditions that affect human performance. The Nuclear Risk Research Center (NRRC) of the Central Research Institute of Electric Power Industry (CRIEPI) has developed an HRA guide that compiles qualitative analysis methods for collecting plant-specific and scenario-specific conditions that affect human performance into "narratives," reflecting the latest research trends, and models for analysis of tasks under extreme conditions.

• Journal of the Korean Mathematical Society
• /
• v.58 no.1
• /
• pp.45-67
• /
• 2021

In this paper, a mathematical dynamical system involving both deterministic (with or without delay) and stochastic "SIR" epidemic model with nonlinear incidence rate in a continuous reactor is considered. A profound qualitative analysis is given. It is proved that, for both deterministic models, if ��d > 1, then the endemic equilibrium is globally asymptotically stable. However, if ��d ≤ 1, then the disease-free equilibrium is globally asymptotically stable. Concerning the stochastic model, the Feller's test combined with the canonical probability method were used in order to conclude on the long-time dynamics of the stochastic model. The results improve and extend the results obtained for the deterministic model in its both forms. It is proved that if ��s > 1, the disease is stochastically permanent with full probability. However, if ��s ≤ 1, then the disease dies out with full probability. Finally, some numerical tests are done in order to validate the obtained results.

• Journal of KIISE:Software and Applications
• /
• v.30 no.3_4
• /
• pp.358-371
• /
• 2003

Automatic word spacing is a process of deciding correct boundaries between words in a sentence including spacing errors. It is very important to increase the readability and to communicate the accurate meaning of text to the reader. The previous statistical approaches for automatic word spacing do not consider the previous spacing state, and thus can not help estimating inaccurate probabilities. In this paper, we propose two statistical word spacing models which can solve the problem of the previous statistical approaches. The proposed models are based on the observation that the automatic word spacing is regarded as a classification problem such as the POS tagging. The models can consider broader context and estimate more accurate probabilities by generalizing hidden Markov models. We have experimented the proposed models under a wide range of experimental conditions in order to compare them with the current state of the art, and also provided detailed error analysis of our models. The experimental results show that the proposed models have a syllable-unit accuracy of 98.33% and Eojeol-unit precision of 93.06% by the evaluation method considering compound nouns.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.38 no.4
• /
• pp.117-131
• /
• 2015

Conventional data envelopment analysis (DEA) models require that inputs and outputs are given as crisp values. Very often, however, some of inputs and outputs are given as imprecise data where they are only known to lie within bounded intervals. While a typical approach to addressing this situation for optimization models such as DEA is to conduct sensitivity analysis, it provides only a limited ex-post measure against the data imprecision. Robust optimization provides a more effective ex-ante measure where the data imprecision is directly incorporated into the model. This study aims to apply robust optimization approach to DEA models with imprecise data. Based upon a recently developed robust optimization framework which allows a flexible adjustment of the level of conservatism, we propose two robust optimization DEA model formulations with imprecise data; multiplier and envelopment models. We demonstrate that the two models consider different risks regarding imprecise efficiency scores, and that the existing DEA models with imprecise data are special cases of the proposed models. We show that the robust optimization for the multiplier DEA model considers the risk that estimated efficiency scores exceed true values, while the one for the envelopment DEA model deals with the risk that estimated efficiency scores fall short of true values. We also show that efficiency scores stratified in terms of probabilistic bounds of constraint violations can be obtained from the proposed models. We finally illustrate the proposed approach using a sample data set and show how the results can be used for ranking DMUs.

• Journal of the Korean Society of Safety
• /
• v.35 no.1
• /
• pp.97-106
• /
• 2020

The objective of this paper is to suggest a new quantification method for multi-unit probabilistic safety assessment (PSA) that removes the overestimation error caused by the existing delete-term approximation (DTA) based quantification method. So far, for the actual plant PSA model quantification, a fault tree with negates have been solved by the DTA method. It is well known that the DTA method induces overestimated core damage frequency (CDF) of nuclear power plant (NPP). If a PSA fault tree has negates and non-rare events, the overestimation in CDF drastically increases. Since multi-unit seismic PSA model has plant level negates and many non-rare events in the fault tree, it should be very carefully quantified in order to avoid CDF overestimation. Multi-unit PSA fault tree has normal gates and negates that represent each NPP status. The NPP status means core damage or non-core damage state of individual NPPs. The non-core damage state of a NPP is modeled in the fault tree by using a negate (a NOT gate). Authors reviewed and compared (1) quantification methods that generate exact or approximate Boolean solutions from a fault tree, (2) DTA method generating approximate Boolean solution by solving negates in a fault tree, and (3) probability calculation methods from the Boolean solutions generated by exact quantification methods or DTA method. Based on the review and comparison, a new intersection removal by probability (IRBP) method is suggested in this study for the multi-unit PSA. If the IRBP method is adopted, multi-unit PSA fault tree can be quantified without the overestimation error that is caused by the direct application of DTA method. That is, the extremely overestimated CDF can be avoided and accurate CDF can be calculated by using the IRBP method. The accuracy of the IRBP method was validated by simple multi-unit PSA models. The necessity of the IRBP method was demonstrated by the actual plant multi-unit seismic PSA models.