• Title/Summary/Keyword: 이중추출

Search Result 601, Processing Time 0.026 seconds

SNS news Recommendation by Using Cosine Similarity (코사인 유사도 기법을 이용한 뉴스 추천 시스템)

  • Kim, Simon;Kim, Hyung-Jun;Han, In-Kyu
    • Annual Conference on Human and Language Technology
    • /
    • 2013.10a
    • /
    • pp.163-166
    • /
    • 2013
  • 사용자별로 SNS/RSS 구독 뉴스 분석을 통해 사용자가 관심이 있는 새로운 뉴스를 추천해 주는 시스템을 설계하고 구현한다. 뉴스 추천 시스템의 설계를 위해 전체 시스템에서 사용자와 서버에서의 작업을 명세하고, 이중에 주요 기능을 담당하는 부분을 구현한다. 구현된 주요 기능은 선호 문서가 들어왔을 때 특징을 추출하고 이를 저장하는 것과 새로운 문서가 들어왔을 때 선호 문서군과 얼마나 유사한지 판별하여 문서에 대한 추천 여부를 결정하는 것이다. 선호 문서의 특징 추출에 대해서는 형태소 분석을 통해 단어와 빈도를 추출하고 이를 누적하여 저장한다. 또한, 새로운 문서가 들어왔을 때 코사인 유사도를 계산하여 사용자가 선호하는 학습문서와의 유사도 비교를 통해 문서 추천 여부를 결정한다. 구현된 시스템에서 실제로 연관된 선호 문서군을 학습시키고, 연관된 새로운 문서 혹은 연관되지 않은 새로운 문서에 대한 추천 여부를 비교하는 것으로 시스템 정확도를 파악한다.

  • PDF

Korean Text Automatic Summarization using Semantically Expanded Sentence Similarity (의미적으로 확장된 문장 간 유사도를 이용한 한국어 텍스트 자동 요약)

  • Kim, Heechan;Lee, Soowon
    • Proceedings of the Korea Information Processing Society Conference
    • /
    • 2014.11a
    • /
    • pp.841-844
    • /
    • 2014
  • 텍스트 자동 요약은 수많은 텍스트 데이터를 처리함에 있어 중요한 연구 분야이다. 이중 추출요약은 현재 가장 많이 연구가 되고 있는 자동 요약 분야이다. 본 논문은 추출 요약의 선두 연구인 TextRank는 문장 간 유사도를 계산할 때 문장 내 단어 간의 의미적 유사성을 충분히 고려하지 못하였다. 본 연구에서는 의미적 유사성을 고려한 새로운 단어 간 유사도 측정 방법을 제안한다. 추출된 문장 간 유사도는 그래프로 표현되며, TextRank의 랭킹 알고리즘과 동일한 랭킹 알고리즘을 사용하여 실험적으로 평가하였다. 그 결과 문장 간 유사성을 고려할 때 단어의 의미적 요소를 충분히 고려하여 정보의 유실을 최소화하여야 한다는 것을 실험 결과로써 확인할 수 있었다.

Effects of Some Herbal Extracts on Aflatoxin $B_1$ Production from Aspergillus parasiticus (생약추출물이 Aspergillus parasiticus Aflatoxin $B_1$ 생성에 미치는 영향)

  • Chung, Sang-Jin
    • Korean Journal of Medicinal Crop Science
    • /
    • v.11 no.5
    • /
    • pp.385-391
    • /
    • 2003
  • The influences of the extracts from Cinnamomi Cortex, Eucommiae Cortex, Puerariae Radix, Lycii Fructus, Zizyphi Fructus, Schisandrae Fructus, Mume Fructus, Chaenomelis Fructus on mycelial growth and aflatoxin $B_1$ production from Aspergillus parasiticus were analyzed. The pH of the culture media were reduced to below pH 4 by all the herbal extracts after 3 days incubation. However, the pH of the culture media increased above pH 6 after 6 days incubation using the extracts from Cinnamomi Cortex, Eucommiae Cortex, Puerariae Radix and Lycii Fructus. The mycelial growth of A. parasiticus was increased over the amount of the control. Puerariae Radix produced the largest amount of mycelial growth and Chaenomelis Fructus produced the smallest amount of mycelial growth. The productions of aflatoxin $B_1$ from A. parasticus culture were increased by the extracts of Puerariae Radix and Zizyphi Fructus, while inhibited by the extracts of Cinnamomi Cortex, Eucommiae Cortex, Lycii Fructus, Schisandrae Fructus, Mume Fructus and Chaenomelis Fructus. In particular, the extracts of Cinnamomi Cortex, Lycii Fructus and Schisandrae Fructus almost inhibited the production of aflatoxin $B_1$. The production of the total protein from Cinnamomi Cortex, which produced much less aflatoxin $B_1$, and Puerariae Radix, which produced a great deal of aflatoxin $B_1$ from A parasticus were slightly higher than the production of the total protein of the control medium.

Enhancing Performance of Bilingual Lexicon Extraction through Refinement of Pivot-Context Vectors (중간언어 문맥벡터의 정제를 통한 이중언어 사전 구축의 성능개선)

  • Kwon, Hong-Seok;Seo, Hyung-Won;Kim, Jae-Hoon
    • Journal of KIISE:Software and Applications
    • /
    • v.41 no.7
    • /
    • pp.492-500
    • /
    • 2014
  • This paper presents the performance enhancement of automatic bilingual lexicon extraction by using refinement of pivot-context vectors under the standard pivot-based approach, which is very effective method for less-resource language pairs. In this paper, we gradually improve the performance through two different refinements of pivot-context vectors: One is to filter out unhelpful elements of the pivot-context vectors and to revise the values of the vectors through bidirectional translation probabilities estimated by Anymalign and another one is to remove non-noun elements from the original vectors. In this paper, experiments have been conducted on two different language pairs that are bi-directional Korean-Spanish and Korean-French, respectively. The experimental results have demonstrated that our method for high-frequency words shows at least 48.5% at the top 1 and up to 88.5% at the top 20 and for the low-frequency words at least 43.3% at the top 1 and up to 48.9% at the top 20.

Depth Generation using Bifocal Stereo Camera System for Autonomous Driving (자율주행을 위한 이중초점 스테레오 카메라 시스템을 이용한 깊이 영상 생성 방법)

  • Lee, Eun-Kyung
    • The Journal of the Korea institute of electronic communication sciences
    • /
    • v.16 no.6
    • /
    • pp.1311-1316
    • /
    • 2021
  • In this paper, we present a bifocal stereo camera system combining two cameras with different focal length cameras to generate stereoscopic image and their corresponding depth map. In order to obtain the depth data using the bifocal stereo camera system, we perform camera calibration to extract internal and external camera parameters for each camera. We calculate a common image plane and perform a image rectification for generating the depth map using camera parameters of bifocal stereo camera. Finally we use a SGM(Semi-global matching) algorithm to generate the depth map in this paper. The proposed bifocal stereo camera system can performs not only their own functions but also generates distance information about vehicles, pedestrians, and obstacles in the current driving environment. This made it possible to design safer autonomous vehicles.

Kalman filter modeling for the estimation of tropospheric and ionospheric delays from the GPS network (망기반 대류 및 전리층 지연 추출을 위한 칼만필터 모델링)

  • Hong, Chang-Ki
    • Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
    • /
    • v.30 no.6_1
    • /
    • pp.575-581
    • /
    • 2012
  • In general, various modeling and estimation techniques have been proposed to extract the tropospheric and ionospheric delays from the GPS CORS. In this study, Kalman filter approach is adopted to estimate the tropospheric and ionospheric delays and the proper modeling for the state vector and the variance-covariance matrix for the process noises are performed. The coordinates of reference stations and the zenith wet delays are estimated with the assumption of random walk stochastic process. Also, the first-order Gauss-Markov stochastic process is applied to compute the ionospheric effects. For the evaluation of the proposed modeling technique, Kalman filter algorithm is implemented and the numerical test is performed with the CORS data. The results show that the atmospheric effects can be estimated successfully and, as a consequence, can be used for the generation of VRS data.

Detection of Mass on Dense Mammogram (고밀도 유방영상에서 종양의 추출)

  • Yu, Seung-Hwa;No, Seung-Mu;Park, Jong-Won
    • Journal of the Institute of Electronics Engineers of Korea SP
    • /
    • v.38 no.6
    • /
    • pp.721-734
    • /
    • 2001
  • This paper proposed automated methods for the detection of breast mass. We analysed characteristic of the mass by using the features on mammograms. The homogeneity was used to distinguish mass and abnormal homogeneous tissue from the Cooper's ligament and multiple threshold method was used to deal with the high density candidates. By using the 8-connectivity, the first step candidates were selected. We generated the dualistic images of each candidate in which we regard the gray value as topographic height information. From these candidates, the second candidates were selected by comparing the circularity and the distribution rates. The final detection was done with the method in which we generated the template of each candidate and compared each other. From these methods, we grade the order from the candidate. We applied the algorithm to the 136 mammograms and compared to the radiologist's outlines of the leisions. The detection resulted that the sensitivity of the proposed methods was 93.38% and 97.63% FP(False positive) which we can segmented mass in the first grade in the 124 cases.

  • PDF

Inhibitory Effects of Garlic Extracts on the Nitrosation (마늘 추출물이 Nitrosamine 생성반응에 미치는 영향)

  • Im, Kyung-Jae;Park, Dong-Ki;Rhee, Moon-Soo;Lee, Joong-Keun;Lee, Si-Kyung
    • Applied Biological Chemistry
    • /
    • v.43 no.2
    • /
    • pp.110-115
    • /
    • 2000
  • This study was carried out to investigate the effect of extracts from garlic on the formation of nitrosamine, nitrosomorphorine, the nitrite scavenging ability and the antioxidant activity. The extracts from garlic made the nitrite scavenging ability higher at low pH than at high pH. Especially ethanol extracts from garlic resulted in 89.6% of nitrite scavenging ability at pH 1.2. They were also more effective in the electron donating and antioxidant abilities than water extracts in expressing as the reduction of ${\alpha},{\alpha}-diphenyl-{\beta}-picrylhydrazyl$(DPPH) and peroxide value(POV), respectively. Inhibitory effect of nitrosamine formation by ethanol extract from garlic was excellently more than 80% at pH 1.2 and 3.0, but only 29.9% at pH 6.0. This meant that the inhibition of nitrosamine formation is pH dependent.

  • PDF

The study on target recognition method to process real-time in W-band mmWave small radar (밀리미터파대역(W-대역)공대지 레이다의 이중편파 채널을 활용한 지상 표적 식별 기법에 관한 연구)

  • Park, Sungho;Kong, Young-Joo;Ryu, Seong-Hyun;Yoon, Jong-Suk
    • The Journal of the Institute of Internet, Broadcasting and Communication
    • /
    • v.18 no.3
    • /
    • pp.61-69
    • /
    • 2018
  • In this paper, we propose a method for recognizing ground target using dual polarization channels in millimeter waveband air-to-surface radar. First, the Push-Broom target detection method is described and the received signal is modeled considering flight-path scenario of air-to-surface radar. The scattering centers were extracted using the RELAX algorithm, which is a time domain spectral estimation technique, and the feature vector of the target was generated. Based on this, a DB for 4 targets is constructed. As a result of the proposed method, it is confirmed that the target classification rates is improved by more than 15% than the single channel using the data of the dual polarization channel.

Design of the Non-Resonant SWG Antenna with Double Slots in the Narrow Wall of Rectangular Waveguide (구형 도파관의 협벽에 이중 슬롯을 가진 비공진형 슬롯 도파관 안테나의 설계)

  • Hur, Moon-Man
    • The Journal of Korean Institute of Electromagnetic Engineering and Science
    • /
    • v.22 no.1
    • /
    • pp.106-113
    • /
    • 2011
  • In this paper, the non-resonant SWG(Slotted Waveguide) antenna with double slots in narrow wall of rectangular waveguide is designed. Because energy radiated from each slot depends upon inclination angle of slot of the designed antenna, inclination angle of each slot is controlled to satisfy the amplitude distribution for required sidelobe level. Instead of the conventional extraction method of slot conductance, this amplitude distribution is made by the proposed method, which employs far-field radiation pattern calculated by Fourier transform of aperture field distribution on slot. The non-resonant double SWG antenna is designed by the proposed method and is manufactured. The antenna performances are measured and compared with the simulated results.