• Proceedings of the Korean Information Science Society Conference
• /
• 2002.04b
• /
• pp.280-282
• /
• 2002

본 논문은 웹 문서 여과시 사용자 모델링을 위해 사용되는 연관성 피드백 방법을 강화 학습 프레임웍에서 분석하고 강화학습 기반의 새로운 연관성 피드백 알고리즘을 제안한다. 제안된 방법은 강화 학습 프레임책상에서 기존의 방법을 일반화한 것으로 기존의 연관성 피드백 방법이 현재의 프로파일만을 상태로 사용하는 데 비해 과거 history부터 얻는 추가 정보를 사용하는 방법이다

• Journal of the Korean Society for information Management
• /
• v.15 no.1
• /
• pp.129-148
• /
• 1998

The objective of this study is to develop a useful relevance feedback retrieval technique that can be applied to the current Boolean retrieval system. A feedback retrieval technique based on user model is recommended here to achieve this objective. To prove the usefulness of this feedback retrieval technique, two enhanced Boolean retrieval models including DNF model and P-norm model were evaluated first through retrieval effectiveness experiments. After selecting DNF model as the retrieval model, two feedback retrieval experiments were performed using initial and extended user models. It is proved that the feedback retrieval based on user model can greatly enhance the effectiveness of a Boolean retrieval system with a small modification.

• 한국HCI학회:학술대회논문집
• /
• 2007.02a
• /
• pp.174-181
• /
• 2007

정보 검색의 결과로 나타나는 요약문을 스니펫(snippet)이라 한다. 사용자는 자신이 원하는 정보를 얻기 위해 문서를 검색하는데, 이 때 스니펫은 사용자가 원하는 문서를 찾는데 중요한 역할을 한다. 본 논문에서는 정보검색 분야에서 높은 성능을 보이는 유사 적합성 피드백을 자동 문서 요약에 맞게 적용하여 높은 성능의 스니펫 생성 시스템을 구현한다. 우선, 사용자의 질의가 포함된 문장들을 일차적으로 요약 문장 후보로 추출한다. 그리고 추출된 문장 후보로부터 명사들을 질의 후보로 고려한다. 각 문장이 질의의 포함 여부에 따라 문장의 적합성을 판단하게 되고, 유사 적합성 피드백 확률 모델에 적용한 후 질의 후보들의 가중치를 추정하여 가중치 순위를 통해 확장할 질의들을 결정한다. 확장된 질의들과 기존의 질의들의 가중치를 합산하여 각 문장의 순위를 매기게 되고 가장 높은 순위의 문장들이 스니펫으로 제시된다. 논문에서 제안한 기법은 추가적인 핵심 질의들을 자동으로 확장하여 중요한 문장을 추출할 수 있다. 이 연구를 위해서 일반 상용 정보 검색 서비스에서 제공하는 스니펫을 수집하였고 이들의 정확도와 시스템의 정확도를 비교하였다. 실험 결과를 통해 살펴본 제안된 시스템의 성능은 상용 정보 검색기에서 제공되고 잇는 스니펫의 정확도 보다 우수한 성능을 보였다.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.39 no.4
• /
• pp.28-37
• /
• 2002

In this paper, we present an image retrieval method for improving retrieval performance by fusion of probabilistic method and query point movement. In the proposed algorithm, the similarity for probabilistic method and the similarity for query point movement are fused in the computation of the similarity between a query image and database image. The probabilistic method used in this paper is suitable for handling negative examples. On the other hand, query point movement deals with the statistical property of positive examples. Combining these two methods, our goal is to overcome their shortcoming. Experimental results show that the proposed method yields better performances over the probabilistic method and query point movement, respectively.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2012.10a
• /
• pp.387-388
• /
• 2012

In order to solve problem which User sometime visits the wrong page with respect to user intention when uses snippet. this paper proposes a new snippet extraction method using fuzzy. The proposed method uses pseudo relevance feedback to expand the use's query. It uses the fuzzy association between the expanded query and the web pages to extract snippet to be well reflected semantic user's intention.

• Journal of Information Processing Systems
• /
• v.9 no.3
• /
• pp.349-364
• /
• 2013

The big challenge in current content-based image retrieval systems is to reduce the semantic gap between the low level-features and high-level concepts. In this paper, we have proposed a novel framework for efficient image retrieval to improve the retrieval results significantly as a means to addressing this problem. In our proposed method, we first extracted a strong set of image features by using the dual-tree rotated complex wavelet filters (DT-RCWF) and dual tree-complex wavelet transform (DT-CWT) jointly, which obtains features in 12 different directions. Second, we presented a relevance feedback (RF) framework for efficient image retrieval by employing a support vector machine (SVM), which learns the semantic relationship among images using the knowledge, based on the user interaction. Extensive experiments show that there is a significant improvement in retrieval performance with the proposed method using SVMRF compared with the retrieval performance without RF. The proposed method improves retrieval performance from 78.5% to 92.29% on the texture database in terms of retrieval accuracy and from 57.20% to 94.2% on the Corel image database, in terms of precision in a much lower number of iterations.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.11
• /
• pp.2374-2381
• /
• 2012

This paper proposes a new enhancing snippet extraction method using fuzzy and semantic features. The proposed method creates a delegate of sentence by using semantic features. It extracts snippet using fuzzy association between a delegate sentence and sentence set which well represents query. In addition, the method uses pseudo relevance feedback to expand query which extracts snippet to be well reflected semantic user's intention. The experimental results demonstrate the proposed method can achieve better snippet extraction performance than the previous methods.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.12 no.1
• /
• pp.15-19
• /
• 2012

A Domain Dictionary generation algorithm based on pseudo feedback model is presented in this paper. This algorithm can increase the precision of domain dictionary generation algorithm. The generation of Domain Dictionary is regarded as a domain term retrieval process: Assume that top N strings in the original retrieval result set are relevant to C, append these strings into the dictionary, retrieval again. Iterate the process until a predefined number of domain terms have been generated. Experiments upon corpus show that the precision of pseudo feedback model based algorithm is much higher than existing algorithms.

• Journal of KIISE:Databases
• /
• v.27 no.1
• /
• pp.90-100
• /
• 2000

We propose, in this paper, a new query expansion technique with term reweighting. All terms in the documents feedbacked from a user, excluding stopwords, are selected as candidate terms for query expansion and reweighted using the relevance degree which is calculated from the term-distribution similarity between a candidate term and each term in initial query. The term-distribution similarity of two terms is a measure on how similar their occurrence distributions in relevant documents are. The terms to be actually expanded are selected using the relevance degree and combined with initial query to construct an expanded query. We use KT-set 1.0 and KT-set 2.0 to evaluate performance and compare our method with two methods, one with no relevance feedback and the other with Dec-Hi method which is similar to our method. based on recall and precision.

• Journal of KIISE:Databases
• /
• v.33 no.4
• /
• pp.372-383
• /
• 2006

Among the approaches to image database management, content-based image retrieval (CBIR) is viewed as having the best support for effective searching and browsing of large digital image libraries. Typical CBIR systems allow a user to provide a query image, from which low-level features are extracted and used to find 'similar' images in a database. However, there exists the semantic gap between human visual perception and low-level representations. An effective methodology for overcoming this semantic gap involves relevance feedback to perform feature re-weighting. Current approaches to feature re-weighting require the number of components for a feature representation to be the same for every image in consideration. Following this assumption, they map each component to an axis in the n-dimensional space, which we call the metric space; likewise the feature representation is stored in a fixed-length vector. However, with the emergence of features that do not have a fixed number of components in their representation, existing feature re-weighting approaches are invalidated. In this paper we propose a feature re-weighting technique that supports features regardless of whether or not they can be mapped into a metric space. Our approach analyses the feature distances calculated between the query image and the images in the database. Two-sided confidence intervals are used with the distances to obtain the information for feature re-weighting. There is no restriction on how the distances are calculated for each feature. This provides freedom for how feature representations are structured, i.e. there is no requirement for features to be represented in fixed-length vectors or metric space. Our experimental results show the effectiveness of our approach and in a comparison with other work, we can see how it outperforms previous work.