• Title/Summary/Keyword: Actual Usage

Search Result 542, Processing Time 0.029 seconds

Fine-tuning BERT-based NLP Models for Sentiment Analysis of Korean Reviews: Optimizing the sequence length (BERT 기반 자연어처리 모델의 미세 조정을 통한 한국어 리뷰 감성 분석: 입력 시퀀스 길이 최적화)

  • Sunga Hwang;Seyeon Park;Beakcheol Jang
    • Journal of Internet Computing and Services
    • /
    • v.25 no.4
    • /
    • pp.47-56
    • /
    • 2024
  • This paper proposes a method for fine-tuning BERT-based natural language processing models to perform sentiment analysis on Korean review data. By varying the input sequence length during this process and comparing the performance, we aim to explore the optimal performance according to the input sequence length. For this purpose, text review data collected from the clothing shopping platform M was utilized. Through web scraping, review data was collected. During the data preprocessing stage, positive and negative satisfaction scores were recalibrated to improve the accuracy of the analysis. Specifically, the GPT-4 API was used to reset the labels to reflect the actual sentiment of the review texts, and data imbalance issues were addressed by adjusting the data to 6:4 ratio. The reviews on the clothing shopping platform averaged about 12 tokens in length, and to provide the optimal model suitable for this, five BERT-based pre-trained models were used in the modeling stage, focusing on input sequence length and memory usage for performance comparison. The experimental results indicated that an input sequence length of 64 generally exhibited the most appropriate performance and memory usage. In particular, the KcELECTRA model showed optimal performance and memory usage at an input sequence length of 64, achieving higher than 92% accuracy and reliability in sentiment analysis of Korean review data. Furthermore, by utilizing BERTopic, we provide a Korean review sentiment analysis process that classifies new incoming review data by category and extracts sentiment scores for each category using the final constructed model.

Temperature Prediction and Control of Cement Preheater Using Alternative Fuels (대체연료를 사용하는 시멘트 예열실 온도 예측 제어)

  • Baasan-Ochir Baljinnyam;Yerim Lee;Boseon Yoo;Jaesik Choi
    • Resources Recycling
    • /
    • v.33 no.4
    • /
    • pp.3-14
    • /
    • 2024
  • The preheating and calcination processes in cement manufacturing, which are crucial for producing the cement intermediate product clinker, require a substantial quantity of fossil fuels to generate high-temperature thermal energy. However, owing to the ever-increasing severity of environmental pollution, considerable efforts are being made to reduce carbon emissions from fossil fuels in the cement industry. Several preliminary studies have focused on increasing the usage of alternative fuels like refuse-derived fuel (RDF). Alternative fuels offer several advantages, such as reduced carbon emissions, mitigated generation of nitrogen oxides, and incineration in preheaters and kilns instead of landfilling. However, owing to the diverse compositions of alternative fuels, estimating their calorific value is challenging. This makes it difficult to regulate the preheater stability, thereby limiting the usage of alternative fuels. Therefore, in this study, a model based on deep neural networks is developed to accurately predict the preheater temperature and propose optimal fuel input quantities using explainable artificial intelligence. Utilizing the proposed model in actual preheating process sites resulted in a 5% reduction in fossil fuel usage, 5%p increase in the substitution rate with alternative fuels, and 35% reduction in preheater temperature fluctuations.

A Study on the Change of Opticion's perception and Actual conditions on Eyewear Dispensing (안경조정에 대한 안경사의 인식 및 실태 변화에 관한 연구)

  • Park, Jeong-Sik;Lee, Jeong-Yeong;Jang, Woo-Yeong
    • Journal of the Korea society of information convergence
    • /
    • v.7 no.1
    • /
    • pp.47-53
    • /
    • 2014
  • This study was carried out recognition and status changes of optician for glasses fitting. It based on a survey results of twice in relation to the glasses fitting. The results are as follows. The ratio of glasses fitting about optician operation appeared to increase gradually. The positive awareness change appeared from the customer and optician about the cost charging about the glasses fitting. In the case of a plastic glasses, the burden for the inconvenience feeling and fitting frequency showed an increase with relation to the temple fitting. To reduce the burden of glasses fitting, The continuous education for the correct glasses usage need.

  • PDF

Multihop Rate Adaptive Wireless Scalable Video Using Syndrome-Based Partial Decoding

  • Cho, Yong-Ju;Radha, Hayder;Seo, Jeong-Il;Kang, Jung-Won;Hong, Jin-Woo
    • ETRI Journal
    • /
    • v.32 no.2
    • /
    • pp.273-280
    • /
    • 2010
  • The overall channel capacity of a multihop wireless path drops progressively over each hop due to the cascading effect of noise and interference. Hence, without optimal rate adaptation, the video quality is expected to degrade significantly at any client located at a far-edge of an ad-hoc network. To overcome this limitation, decoding and forwarding (DF), which fully decodes codewords at each intermediate node, can be employed to provide the best video quality. However, complexity and memory usage for DF are significantly high. Consequently, we propose syndrome-based partial decoding (SPD). In the SPD framework an intermediate node partially decodes a codeword and relays the packet along with its syndromes if the packet is corrupted. We demonstrate the efficacy of the proposed scheme by simulations using actual 802.11b wireless traces. The trace-driven simulations show that the proposed SPD framework, which reduces the overall processing requirements of intermediate nodes, provides reasonably high goodput when compared to simple forwarding and less complexity and memory requirements when compared to DF.

Using a Modular BOM for Cost Calculation in Wiring Harness (Modular BOM을 활용한 와이어링 하네스 원가 산출)

  • Im, Chang-Sung;Park, Hoon;Kang, Jung-Tae
    • Korean Journal of Computational Design and Engineering
    • /
    • v.21 no.1
    • /
    • pp.61-68
    • /
    • 2016
  • Since past till now, automotive wiring harness has been designed and produced under the method of part number list. However, the number of options has increased gradually on the basis of customer's needs and car manufacturer's policy, thereby bringing about an increase in the number of classified parts. Because of a large number of classified parts, cost estimates for manufacturing and production planning are more important than ever. However, the pre-existing manufacturing bill of material abbreviated as BOM is only focused on the list of materials used for actual manufacturing, and as such there have been limits to other fields or areas of usage except for manufacturing. In essence, the present paper is concerned with our study conducted on modular BOM as one of planning BOM. On such basis, we have been able to completely use a method by which to have access, make an option units' calculation in wiring harness manufacturing, and build an efficient system for cost estimates in manufacturing as well.

A Study on the Characteristics of Naturalism in Fashion Design with the Changes of Times (자연주의 복식의 시대적 변천에 따른 디자인 특성에 관한 연구)

  • Kim Hee-Yeon;Kim Young-In
    • Journal of the Korean Society of Costume
    • /
    • v.56 no.7 s.106
    • /
    • pp.31-41
    • /
    • 2006
  • The purpose of this study is to formulate a concept of naturalism in fashion and verify the design characteristics found in such designs via reviewing the concept of naturalism and its influence on dress making in different periods in history. The references from various academic fields including philosophy, literature and fine arts were used to examine how the naturalism was formulated over the years. This study was then to identify the naturalistic influences by closely examining various Western dress codes found in literature, publications and actual samples. The results of this study shows that the close examination of naturalism in fashion with the changes of times displayed the complex evolution process over the years, and they were classified by four distinctive categories; Retro, Ecology, Primitive and Ethnic. Naturalism in the contemporary fashion design showed higher frequency of adopting ecological and retro styles, especially retro style adoption being on the rise, combining wide range of partial components of each style. The color usage showed the skewed concentration on YellowRed, Red and Yellow colors with pale, light grayish and grayish tones.

A Study of Unaccounted-for Gas Volume : Seoul, Kyunggi Area (미설명 도시가스의 분석 : 경인지역 중심으로)

  • 김수덕;서정규
    • Journal of Energy Engineering
    • /
    • v.12 no.4
    • /
    • pp.328-338
    • /
    • 2003
  • A study is perfomed on the difference between the volume of gas purchased from Korea Gas Corporation (KGC) and that of measured gas deliveries including total sales, pipeline inventory, and other usage by gas companies at Seoul and Kyunggi area. Actual data md information from 1996 to 2001 on gas purchase from KGC, sales, loss and own company use and pipeline status of each gas company are used to identify net inventory change in pipeline, unaccounted inventory estimates based on billing cycle and their net changes. A framework to be used for the analysis of such differences and an empirical test result are being provided. As a conclusion, the hypothesis that each gas company purchased more than the volume of gas accounted-for by all the possible factors is shown to be rejected.

A Study on the Magnetic Field Analysis and Optimal Core Design of DC Current Sensor for Vehicles (자동차용 DC Current Sensor의 자장해석 및 코어 최적형상 설계에 관한 연구)

  • Lee, Hee-Sung;Park, Jong-Min;Kim, Choon-Sik;Kim, Sung-Gaun
    • Transactions of the Korean Society of Automotive Engineers
    • /
    • v.17 no.5
    • /
    • pp.74-83
    • /
    • 2009
  • Recently, usage of electric and electronic system for car increases rapidly. Consequently power monitoring supplied to the system is essential for management and controlling. Generally, battery status is monitored through measuring and diagnosing the current measurement method utilizing Hall Effect. Therefore, in this paper, we analysed magnetic field to develop the solution of DC current sensor using Hall Effect which is the core of design and development. By analysing the magnetic field by FEM using Maxwell 3D software, the location of the highest output current and stable part in the Hall IC sensor was shown. Also, the optimal core design of DC current sensor using parametric and Simplex method was presented. A car battery charge and discharge process dependant on time effect on the changing of magnetic field was simulated and compared to the result from the experiment result of actual vehicle.

Using Omnidirectional Images for Semi-Automatically Generating IndoorGML Data

  • Claridades, Alexis Richard;Lee, Jiyeong;Blanco, Ariel
    • Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
    • /
    • v.36 no.5
    • /
    • pp.319-333
    • /
    • 2018
  • As human beings spend more time indoors, and with the growing complexity of indoor spaces, more focus is given to indoor spatial applications and services. 3D topological networks are used for various spatial applications that involve navigation indoors such as emergency evacuation, indoor positioning, and visualization. Manually generating indoor network data is impractical and prone to errors, yet current methods in automation need expensive sensors or datasets that are difficult and expensive to obtain and process. In this research, a methodology for semi-automatically generating a 3D indoor topological model based on IndoorGML (Indoor Geographic Markup Language) is proposed. The concept of Shooting Point is defined to accommodate the usage of omnidirectional images in generating IndoorGML data. Omnidirectional images were captured at selected Shooting Points in the building using a fisheye camera lens and rotator and indoor spaces are then identified using image processing implemented in Python. Relative positions of spaces obtained from CAD (Computer-Assisted Drawing) were used to generate 3D node-relation graphs representing adjacency, connectivity, and accessibility in the study area. Subspacing is performed to more accurately depict large indoor spaces and actual pedestrian movement. Since the images provide very realistic visualization, the topological relationships were used to link them to produce an indoor virtual tour.

Usage and Estimation of R-indicator for Representative (대표성을 위한 R-indicator의 사용과 추정법 연구)

  • Park, Hyeonah;Lee, Kee-Jae
    • The Korean Journal of Applied Statistics
    • /
    • v.28 no.3
    • /
    • pp.417-427
    • /
    • 2015
  • Measures in response rate used to measure the representativeness of the sample (the more high response rate) better explain the representativeness of the sample. However, we cannot often explain the representativeness of the sample because there is nonresponse even in the high response rate. Therefore, Schouten et al. (2009) presented a new R-indicator measure that can be described as a representative of the sample. We research the new estimator of the R-indicator in this paper because there are parameters that require estimations. We describe the meanings as representative of the R-indicator; consequently, the bias and efficiency of the proposed estimator for R-indicator are compared to the existing estimator under various simulations. The representativeness of the sample is also explained by applying the proposed estimators in the actual data.