• Title/Summary/Keyword: and Information Retrieval

검색결과 3,440건 처리시간 0.034초

음성 데이터베이스로부터의 효율적인 색인데이터베이스 구축과 정보검색 (The Extraction of Effective Index Database from Voice Database and Information Retrieval)

  • 박미성
    • 한국도서관정보학회지
    • /
    • 제35권3호
    • /
    • pp.271-291
    • /
    • 2004
  • 전자도서관과 같은 정보제공원은 이미지, 음성, 동영상 등과 같은 비정형 멀티미디어 데이터 서비스에 대한 요구를 받고 있다. 그리하여 본 연구에서는 음성 처리를 위해 어절생성기, 음절복원기, 형태소분석기, 교정기를 제안하였다. 제안한 음성처리 기술로 음성데이터베이스를 텍스트데이터베이스로 변환 한후 텍스트데이터베이스로부터 색인데이터베이스를 추출하였다. 그리고 추출한 색인데이터베이스로 텍스트와 음성의 내용기반정보검색에 활용할 수 있음을 보이기 위해 정보검색모델을 제안하였다.

  • PDF

칼라와 에지 정보를 이용한 내용기반 영상 검색 (Contents-based Image Retrieval Using Color & Edge Information)

  • 박동원;안성옥
    • 컴퓨터교육학회논문지
    • /
    • 제8권1호
    • /
    • pp.81-91
    • /
    • 2005
  • 본 논문에서는 칼라와 에지 정보를 이용한 내용기반 영상검색 기법을 제안하였다. 기존의 RGB 공간 정보를 이용하기 보다는, 시각적 인식에 보다 중점을 둔 HSI칼라 공간에서 고찰하였다. 비슷한 류의 색을 대표색으로 통합 표현하여, 개선된 칼라 정보 이용법을 본 연구에서 제안하였다. 또한 칼라 정보만을 이용했을 때의 시스템 성능상의 결점을 보완하기 위하여, 효율적인 에지 디텍션 기법을 함께 사용하였다. 칼라와 에지 기법을 통합함에 있어서, 각각의 기법에 적절한 가중치를 배분함으로써 시스템 성능을 실험적으로 향상시켰다.

  • PDF

색인어 퍼지 관계와 서열기법을 이용한 정보 검색 방법론 (A Methodology of the Information Retrieval System Using Fuzzy Connection Matrix and Document Connectivity Order)

  • 김철;이승채;김병기
    • 한국정보처리학회논문지
    • /
    • 제3권5호
    • /
    • pp.1160-1169
    • /
    • 1996
  • 본 연구에서는 색인어 퍼지 관계행렬을 이용한 정보검색 방법을 제안하고 간단한 문헌정보 검색시스템을 사용하여 실험을 수행하고 그 결과를 분석하였다. 불리안 연산 자인 AND, OR, NOT으로 색인어들을 조합한 질의식을 통해 실험을 수행한 결과 일반 집합이론에 의한 검색실험에서보다 상당히 우수한 성능을 보였다. 특히 재현율과 정확 률을 측정한 성능평가 결과는 퍼지 문헌검색 시스템이 가능한 검색 대안이라는 사실을 확인 하였다고 할 수 있다. 한편, 검색의 기법 측면에서 고려하였을 때 본 실험은 먼저, 색인어 관계행렬에 따라서 검색결과에 서열을 부여하였고, 기준적합도값의 변동에 따라 검색결과가 유동적으로 대응하도록 하였으며, 관계값을 의미적 거리로 파악함으 로써 검색과정과 검색 시맨틱스를 일치시키고자 새롭게 시도하였다.

  • PDF

Semi-supervised Cross-media Feature Learning via Efficient L2,q Norm

  • Zong, Zhikai;Han, Aili;Gong, Qing
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제13권3호
    • /
    • pp.1403-1417
    • /
    • 2019
  • With the rapid growth of multimedia data, research on cross-media feature learning has significance in many applications, such as multimedia search and recommendation. Existing methods are sensitive to noise and edge information in multimedia data. In this paper, we propose a semi-supervised method for cross-media feature learning by means of $L_{2,q}$ norm to improve the performance of cross-media retrieval, which is more robust and efficient than the previous ones. In our method, noise and edge information have less effect on the results of cross-media retrieval and the dynamic patch information of multimedia data is employed to increase the accuracy of cross-media retrieval. Our method can reduce the interference of noise and edge information and achieve fast convergence. Extensive experiments on the XMedia dataset illustrate that our method has better performance than the state-of-the-art methods.

Acceleration of Building Thesaurus in Fuzzy Information Retrieval Using Relational products

  • Kim, Chang-Min;Kim, Young-Gi
    • 한국지능시스템학회:학술대회논문집
    • /
    • 한국퍼지및지능시스템학회 1998년도 The Third Asian Fuzzy Systems Symposium
    • /
    • pp.240-245
    • /
    • 1998
  • Fuzzy information retrieval which uses the concept of fuzzy relation is able to retrieve documents in the way based on not morphology but semantics, dissimilar to traditional information retrieval theories. Fuzzy information retrieval logically consists of three sets : the set of documents, the set of terms and the set of queries. It maintains a fuzzy relational matrix which describes the relationship between documents and terms and creates a thesaurus with fuzzy relational product. It also provides the user with documents which are relevant to his query. However, there are some problems on building a thesaurus with fuzzy relational product such that it has big time complexity and it uses fuzzy values to be processed with flating-point. Actually, fuzzy values have to be expressed and processed with floating-point. However, floating-point operations have complex logics and make the system be slow. If it is possible to exchange fuzzy values with binary values, we could expect sp eding up building the thesaurus. In addition, binary value expressions require just a bit of memory space, but floating -point expression needs couple of bytes. In this study, we suggest a new method of building a thesaurus, which accelerates the operation of the system by pre-applying an ${\alpha}$-cut. The experiments show the improvement of performance and reliability of the system.

  • PDF

양방향 반올림 필터를 이용한 객체 영역 기반 고속 영상 검색 (Fast Image Retrieval Based on Object Regions Using Bidirectional Round Filter)

  • 류권열;강경원
    • 한국멀티미디어학회논문지
    • /
    • 제6권2호
    • /
    • pp.240-246
    • /
    • 2003
  • 본 논문에서는 웨이브릿 변환 영역에서 양방향 반올림 필터를 이용한 객체 영역 기반 고속 영상 검색 방법을 제안한다. 기존의 방법은 웨이브릿 변환 영역의 부대역 전체에서 특징 벡터를 추출하기 때문에 불필요한 배경 정보가 포함됨으로써 검색 효율이 감소하였다. 제안한 방법은 양방향 반올림 필터를 이용하여 객체 영역에서만 특징 벡터를 추출함으로써 불필요한 배경 정보를 제거하여 검색 효율을 향상시킨다. 그리고, 색상 정보에 관한 특징 벡터 수를 감소하여도 일정한 검색 효율을 유지한다. 결론적으로, 영상의 특성에 따라 다소 차이는 있으나 2.5%∼5.3%의 검색 효율이 향상됨을 알 수 있었다.

  • PDF

Large-Scale Phase Retrieval via Stochastic Reweighted Amplitude Flow

  • Xiao, Zhuolei;Zhang, Yerong;Yang, Jie
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제14권11호
    • /
    • pp.4355-4371
    • /
    • 2020
  • Phase retrieval, recovering a signal from phaseless measurements, is generally considered to be an NP-hard problem. This paper adopts an amplitude-based nonconvex optimization cost function to develop a new stochastic gradient algorithm, named stochastic reweighted phase retrieval (SRPR). SRPR is a stochastic gradient iteration algorithm, which runs in two stages: First, we use a truncated sample stochastic variance reduction algorithm to initialize the objective function. The second stage is the gradient refinement stage, which uses continuous updating of the amplitude-based stochastic weighted gradient algorithm to improve the initial estimate. Because of the stochastic method, each iteration of the two stages of SRPR involves only one equation. Therefore, SRPR is simple, scalable, and fast. Compared with the state-of-the-art phase retrieval algorithm, simulation results show that SRPR has a faster convergence speed and fewer magnitude-only measurements required to reconstruct the signal, under the real- or complex- cases.

Learning Discriminative Fisher Kernel for Image Retrieval

  • Wang, Bin;Li, Xiong;Liu, Yuncai
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제7권3호
    • /
    • pp.522-538
    • /
    • 2013
  • Content based image retrieval has become an increasingly important research topic for its wide application. It is highly challenging when facing to large-scale database with large variance. The retrieval systems rely on a key component, the predefined or learned similarity measures over images. We note that, the similarity measures can be potential improved if the data distribution information is exploited using a more sophisticated way. In this paper, we propose a similarity measure learning approach for image retrieval. The similarity measure, so called Fisher kernel, is derived from the probabilistic distribution of images and is the function over observed data, hidden variable and model parameters, where the hidden variables encode high level information which are powerful in discrimination and are failed to be exploited in previous methods. We further propose a discriminative learning method for the similarity measure, i.e., encouraging the learned similarity to take a large value for a pair of images with the same label and to take a small value for a pair of images with distinct labels. The learned similarity measure, fully exploiting the data distribution, is well adapted to dataset and would improve the retrieval system. We evaluate the proposed method on Corel-1000, Corel5k, Caltech101 and MIRFlickr 25,000 databases. The results show the competitive performance of the proposed method.

Query Formulation for Heuristic Retrieval in Obfuscated and Translated Partially Derived Text

  • Kumar, Aarti;Das, Sujoy
    • Journal of Information Science Theory and Practice
    • /
    • 제3권1호
    • /
    • pp.24-39
    • /
    • 2015
  • Pre-retrieval query formulation is an important step for identifying local text reuse. Local reuse with high obfuscation, paraphrasing, and translation poses a challenge of finding the reused text in a document. In this paper, three pre-retrieval query formulation strategies for heuristic retrieval in case of low obfuscated, high obfuscated, and translated text are studied. The strategies used are (a) Query formulation using proper nouns; (b) Query formulation using unique words (Hapax); and (c) Query formulation using most frequent words. Whereas in case of low and high obfuscation and simulated paraphrasing, keywords with Hapax proved to be slightly more efficient, initial results indicate that the simple strategy of query formulation using proper nouns gives promising results and may prove better in reducing the size of the corpus for post processing, for identifying local text reuse in case of obfuscated and translated text reuse.