• Journal of the Semiconductor & Display Technology
• /
• v.13 no.2
• /
• pp.45-52
• /
• 2014

This paper proposes an efficient photo retrieval system that automatically indexes for searching of relevant images, using a combination of geo-coded information, direction/location of image capture device and content-based visual features. A photo image is labeled with its GPS (Global Positioning System) coordinates and direction of the camera view at the moment of capture, and the label leads to generate a geo-spatial index with three core elements of latitude, longitude and viewing direction. Then, content-based visual features are extracted and combined with the geo-spatial information, for indexing and retrieving the photo images. For user's querying process, the proposed method adopts two steps as a progressive approach, filtering the relevant subset prior to use a content-based ranking function. To evaluate the performance of the proposed scheme, we assess the simulation performance in terms of average precision and F-score, using a natural photo collection. Comparing the proposed approach to retrieve using only visual features, an improvement of 20.8% was observed. The experimental results show that the proposed method exhibited a significant enhancement of around 7.2% in retrieval effectiveness, compared to previous work. These results reveal that a combination of context and content analysis is markedly more efficient and meaningful that using only visual feature for image search.

• Journal of Information Science Theory and Practice
• /
• v.12 no.3
• /
• pp.36-48
• /
• 2024

The information retrieval (IR) process often encounters a challenge known as query-document vocabulary mismatch, where user queries do not align with document content, impacting search effectiveness. Automatic query expansion (AQE) techniques aim to mitigate this issue by augmenting user queries with related terms or synonyms. Word embedding, particularly Word2Vec, has gained prominence for AQE due to its ability to represent words as real-number vectors. However, AQE methods typically expand individual query terms, potentially leading to query drift if not carefully selected. To address this, researchers propose utilizing median vectors derived from deep median networks to capture query similarity comprehensively. Integrating median vectors into candidate term generation and combining them with the BM25 probabilistic model and two IR strategies (EQE1 and V2Q) yields promising results, outperforming baseline methods in experimental settings.

• Asia Marketing Journal
• /
• v.16 no.4
• /
• pp.1-18
• /
• 2015

Assuming that not all background music in advertising function as effective retrieval cues for the advertised messages, this study proposes that we should be able to distinguish the retrieval cue effect of music from the simple ad exposure effect. This study tries to identify which specific characteristics of music are related to the retrieval cue effect. Our experiment focuses on congruency and familiarity of music as key characteristics of music which affect the effectiveness of the music's role as a retrieval cue for the advertised messages. We used four groups of subjects to test the retrieval cue effect of the background music. Each group was exposed to one of the four different types of background music and was again sub-divided into an experimental and a control group (i.e., a total of eight independent sub-groups were included in the experiment.) The first two sub-groups were exposed to the experimental advertisement with the background music of high congruency and high familiarity. After the ad exposure, the background music was provided as a retrieval cue to only one of the two sub-groups. Comparison of the memory performance between the two sub-groups will reveal the net retrieval cue effect of the music of high congruency and high familiarity. Similarly, another two sub-groups watched the same ad but with the background music of high congruency and low familiarity. Also the same ad but with the music of low congruency/ high familiarity was shown to another two sub-groups and that of low congruency and low familiarity music was to another two. Among the two sub-groups with the same music, only one group had the music cue at the memory tasks. One hundred and seventy four undergraduate students at the college of one of authors in Asia participated in the study. Their ages ranged from 18 to 24 with a median of 20. The sample was composed of 51.7 percent male subjects. They were randomly assigned to each of the eight sub-group. The results show that the music highly congruent with the advertised message facilitates the message retrieval, while the low congruency music cue does not. It was also found that the low familiarity music cue improves memory performance only when the music is perceived as congruent with the advertised message. From a theoretical and practical perspective, this study provides boundary conditions for effective retrieval and suggests that the congruent music specifically created for the ad is a more effective retrieval cue than other types of music cues.

• The Transactions of the Korea Information Processing Society
• /
• v.5 no.11
• /
• pp.2776-2785
• /
• 1998

An image retrieval system based on image content is a key issue for building and managing large multimedia database, such as art galleries and museums, trademarks and copyrights, and picture archiving and communication system. Therefore, the interest on the subject of content-based image retrieval has been greatly increased for the last few years. This paper proposes a feature-based image retrieval technique which uses a compound feature vector representing both of color and shape of an image. Color information for the feature vector is obtained using the algebraic moment of each pixel of an image based on the property of regional color distribution. Shape information for the feature vector is obtained using the Improved Moment Invariant(IMI) which reduces the quantity of computation and increases retrieval efficiency. In the preprocessing phase for extracting shape feature, we transform a color image into a gray image. Since we make use of the modified DCT algorithm, it is implemented easily and can extract contour in real time. As an experiment, we have compared our method with previous methods using a database consisting of 150 automobile images, and the results of the experiment have shown that our method has the better performance on retrieval effectiveness.

• Journal of KIISE:Software and Applications
• /
• v.31 no.2
• /
• pp.126-136
• /
• 2004

This paper proposes a new content-based image retrieval algorithm using color-spatial information. For the purpose, the paper suggests two kinds of indexing key to prune away irrelevant images to a given query image; MCS(Main Colors' Set), which is related with color information and DBS (Distribution Block Signature), which is related with spatial information. After successively applying these filters to a database, we could get a small amount of high potential candidates that are somewhat similar to the query image. Then we would make use of new QM(Quad modeling) and relevance feedback mechanism to obtain more accurate retrieval. It would enhance the retrieval effectiveness by dynamically modulating the weights of color-spatial information. Experiments show that the proposed algorithm can apply successfully image retrieval applications.

• Proceedings of the Korean Society for Information Management Conference
• /
• 1996.08a
• /
• pp.15-18
• /
• 1996

본 논문은 본문 전체가 기계가독형 화일로 구성된 텍스트 전문데이터베이스와 이미지화일로 구성된 이미지 전문데이터베이스와의 검색효율성을 비교함으로써 도서관과 최종이용자의 입장에서 바람직한 전문데이터베이스가 어떤 것인지를 제안하고자 한다.

• Proceedings of the IEEK Conference
• /
• 2002.06c
• /
• pp.211-214
• /
• 2002

Automatic video partitioning is the first step for content-based indexing and retrieval of video data. In this paper, an efficient algorithm for scene change detection is proposed, where cumulative histogram and edge information are utilized. Experimental results have shown the effectiveness of the proposed algorithm.

• Journal of the Korean Society for Library and Information Science
• /
• v.41 no.1
• /
• pp.185-214
• /
• 2007

This study intends to design the bibliographic framework based on IFLA-FRBR model for television program materials and to evaluate this in terms of effectiveness of retrieval and usability of the system. The FRBR model supplies mote suitable bibliographic framework of audio-visual material which has a sufficient hierarchical relations and relative bibliographical records. The followings are research methods designed by this study; (1) The experimental metadata system named it FbCS based on FRBR was developed by using the entity-related database and composed of multi-layed and hierarchy. FbCS is developed through benchmarking of a case study for iMMix model in Netherlands based on FRBR. (2) To evaluate effectiveness of retrieval and usability of FbCS, this study made a experiment and survey by user groups of professionals.

• Journal of the Korean Society for Library and Information Science
• /
• v.31 no.1
• /
• pp.153-184
• /
• 1997

The importance of WWW search services is increasing as Internet information resources explode. An evaluation of current 9 search services was first conducted by comparing descriptively the features concerning indexing, searching, and ranking of search results. Secondly, a couple of search queries were used to evaluate search performance of those services by the measures of retrieval effectiveness. the degree of overlap in searching sites, and the degree of similarity between services. In this experiment, Alta Vista, HotBot and Open Text Index showed better results for the retrieval effectiveness. The level of similarity among the 9 search services was extremely low.

• Journal of the Korean Society for Library and Information Science
• /
• v.44 no.4
• /
• pp.75-94
• /
• 2010

This study designed a pilot system in which queries can be expanded through a tag ontology where equivalent, synonymous, or related tags are bound together, in order to improve the retrieval effectiveness of videos. We evaluated the proposed pilot system by comparing it to a tag-based system without tag control, in terms of recall and precision rates. Our study results showed that the mean recall rate in the structured folksonomy-based system was statistically higher than that in the tag-based system. On the other hand, the mean precision rate in the structured folksonomy-based system was not statistically higher than that in the tag-based system. The result of this study can be utilized as a guide on how to effectively use tags as social metadata of digital video libraries.