• Journal of the Korean Society for information Management
• /
• v.14 no.2
• /
• pp.123-141
• /
• 1997

This study examined the extent to which variations in relevance assessments affect the evaluation of the performance of full-text retrieval system. Four sets of relevance judgments obtained by examining the full-text of documents were used to test the retrieval effectiveness. There was no noticeable difference in retrieval performance among the four relevance judgment sets. It implies that a variety of definitions of relevance has no effect on the evaluation of the performance of the full-text retrieval system. Furth r retrieval experiments on this topic incorporating relevance feedback, which is one of the sophisticated retrieval techniques using relevance information, are suggested.

• Annual Conference on Human and Language Technology
• /
• 1997.10a
• /
• pp.65-71
• /
• 1997

이 논문은 부산대학교 전자계산학과 인공지능 연구실에서 개발한 정보검색 시스템 "미리내"의 적합성 피드백 방법을 분석하고, 그 방법들의 검색 효율을 비교 분석하였다. "미리내"에서 질의문은 자연언어 질의문을 사용하고 재검색을 위한 적합성 피드백은 원질의문에서 검색된 문서 중 이용자가 직접 선택한 적합 문서에서 추출한다. 적합성 피드백은 크게 단어 확장(Term Expansion)을 위한 단어 선택 방법과 추가될 단어에 가중치를 부여하는 단어 가중치 부여(Term Weighting)의 2가지 요소로 이루어진다. 단어 선택을 위해서는 적합 문서에 나타난 단어 빈도합(tf), 역문헌빈도(idf), 적합 문서 중에서 해당 단어가 있는 적합 문서의 비율(r/R) 등의 정보를 이용한다. 단어 가중치 부여 방법으로는 정규화 또는 코사인 함수를 이용하여 부여하였다. 단어확장에는 tfidf가 tfidf(r/R)보다 정확도 면에서 나은 향상율을 보였으나, 30위 내 검색된 적합문서의 수를 비교해 보았을 때 tfidf(r/R)의 정확도가 높았다. 단어 선택 방법에서 계산된 값을 정규화하여 가중치를 부여하였을 때 보다 코사인 함수를 이용하여 가중치를 부여하였을 때 정확도가 높았다. 실험은 KT-Set 2.0 (4391 건), 동아일보 96 년 신문기사(70459 건)를 대상으로 수행하였다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.7
• /
• pp.1517-1524
• /
• 2011

In this paper, a new document summarization method, which uses the semantic features and the pseudo relevance feedback (PRF) by using WordNet, is introduced to extract meaningful sentences relevant to a user query. The proposed method can improve the quality of document summaries because the inherent semantic of the documents are well reflected by the semantic feature from NMF. In addition, it uses the PRF by the semantic features and WordNet to reduce the semantic gap between the high level user's requirement and the low level vector representation. The experimental results demonstrate that the proposed method achieves better performance that the other methods.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.5 no.3
• /
• pp.597-604
• /
• 2001

In conventional boolean retrieval systems, document ranking is not supported and similarity coefficients cannot be computed between queries and documents. The MMM, Paice and P-norm models have been proposed in the past to support the ranking facility for boolean retrieval systems. They have common properties of interpreting boolean operators softly. In this paper we propose a new soft evaluation method for Information retrieval using query splitting relevance feedback model. We also show through performance comparison that query splitting relevance feedback(QSRF) is more efficient and effective than MMM, Paice and P-norm.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.683-688
• /
• 2008

The emotional information processing is to simulate and recognize human sensibility, sensuality or emotion, to realize natural and harmonious human-machine interface. This paper proposes an emotion-based image retrieval method. In this method, user can choose a linguistic query among some emotional adjectives. Then the system shows some corresponding representative images that are pre-evaluated by experts. Again the user can select a representative one among the representative images to initiate traditional content-based image retrieval (CBIR). By this proposed method any CBIR can be easily expanded as emotion-based image retrieval. In CBIR of our system, we use several color and texture visual descriptors recommended by MPEG-7. We also propose a fuzzy similarity measure based on Choquet integral in the CBIR system. For the communication between system and user, a relevance feedback mechanism is used to represent human subjectivity in image retrieval. This can improve the performance of image retrieval, and also satisfy the user's individual preference.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.3
• /
• pp.252-257
• /
• 2001

In conventional boolean retrieval systems, document ranking is not supported and similarity coefficients cannot be computed between queries and documents. The MMM, Paice and P-norm models have been proposed in the past to support the ranking facility for boolean retrieval systems. They have common properties of interpreting boolean operators softly. In this paper we propose a new soft evaluation method for Information retrieval using query splitting relevance feedback model. We also show through performance comparison that query splitting relevance feedback(QSRF) is more efficient and effective than MMM, Paice and P-norm.

• Journal of Korean Institute of Industrial Engineers
• /
• v.26 no.4
• /
• pp.363-375
• /
• 2000

In this paper we propose a content-based image retrieval method that can search large image databases efficiently by color, texture, and shape content. Quantized RGB histograms and the dominant triple (hue, saturation, and value), which are extracted from quantized HSV joint histogram in the local image region, are used for representing global/local color information in the image. Entropy and maximum entry from co-occurrence matrices are used for texture information and edge angle histogram is used for representing shape information. Relevance feedback approach, which has coupled proposed features, is used for obtaining better retrieval accuracy. Simulation results illustrate the above method provides 77.5 percent precision rate without relevance feedback and increased precision rate using relevance feedback for overall queries. We also present a new indexing method that supports fast retrieval in large image databases. Tree structures constructed by k-means algorithm, along with the idea of triangle inequality, eliminate candidate images for similarity calculation between query image and each database image. We find that the proposed method reduces calculation up to average 92.9 percent of the images from direct comparison.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.1 s.307
• /
• pp.39-44
• /
• 2006

In this paper, search result list synthesis method is proposed using borda count method for still image retrieval based on MPEG-7 descriptors. MPEG-7 standardizes descriptors that extract feature information from media data. In many cases, using a single descriptor lacks of correctness, it is suggested to use multiple descriptors to enhance retrieval efficiency. In this paper, retrieval efficiency enhancement is achieved by combining multiple search results which are from each descriptor. In combining search result, newly calculated borda count method is proposed. Comparing current frequency compensated calculation, rank considered frequency compensation is used to score animage in database. This combining method is considered in Content based image retrieval system with relevance feedback algorithm which uses high level information from system user. In each relevance iteration step, adoptive borda count method is used to calculate score of images.

• Journal of the Korean Society for Library and Information Science
• /
• v.35 no.3
• /
• pp.173-189
• /
• 2001

This paper presents a framework for observable behavior that can be used as a basis for user modeling, and it reports the results of a pair of user studies that examine the joint utility of two specific behaviors. User models can be constructed by hand, or they can be teamed automatically based on feedback provided by the user about the relevance of documents that they have examined. By observing user behavior, it is possible to obtain implicit feedback without requiring explicit relevance judgments. Four broad categories of potentially observable behavior are identified : examine, retain, reference, and annotate, and examples of specific behaviors within a category are further subdivided based on the natural scope of information objects being manipulated . segment object, or class. Previous studies using Internet discussion groups (USENET news) have shown reading time to be a useful source of implicit feedback for predicting a user's preferences. The experiments reported in this paper extend that work to academic and professional journal articles and abstracts, and explore the relationship between printing behavior and reading time. Two user studies were conducted in which undergraduate students examined articles or abstracts from the telecommunications or pharmaceutical literature. The results showed that reading time can be used to predict the user's assessment of relevance, that the mean reading time for journal articles and technical abstracts is longer than has been reported for USENET news documents, and that printing events provide additional useful evidence about relevance beyond that which can be inferred from reading time. The paper concludes with a brief discussion of the implications of the reported results.

• Journal of the Korean Society for information Management
• /
• v.41 no.1
• /
• pp.261-282
• /
• 2024

This study proposed and evaluated electroencephalography (EEG)-based and eye-tracking-based methods to determine relevance by utilizing users' implicit relevance feedback while navigating content in a digital library. For this, EEG/eye-tracking experiments were conducted on 32 participants using video, image, and text data. To assess the usefulness of the proposed methods, deep learning-based artificial intelligence (AI) techniques were used as a competitive benchmark. The evaluation results showed that EEG component-based methods (av_P600 and f_P3b components) demonstrated high classification accuracy in selecting relevant videos and images (faces/emotions). In contrast, AI-based methods, specifically object recognition and natural language processing, showed high classification accuracy for selecting images (objects) and texts (newspaper articles). Finally, guidelines for implementing a digital library interface based on EEG, eye-tracking, and artificial intelligence technologies have been proposed. Specifically, a system model based on implicit relevance feedback has been presented. Moreover, to enhance classification accuracy, methods suitable for each media type have been suggested, including EEG-based, eye-tracking-based, and AI-based approaches.