• The Journal of the Acoustical Society of Korea
• /
• v.36 no.6
• /
• pp.436-441
• /
• 2017

Computing music similarity is indispensable in constructing music retrieval system. This paper focuses on the cover song search among various music-retrieval tasks. We investigate the cover song search method based on the chroma n-gram to reduce storage for feature DB and enhance search accuracy. Specifically we propose t-tab n-gram, n-gram selection method, and n-gram set comparison method. Experiments on the widely used music dataset confirmed that the proposed method improves cover song search accuracy as well as reduces feature storage.

• The Journal of the Acoustical Society of Korea
• /
• v.37 no.6
• /
• pp.518-524
• /
• 2018

The cover song refers to live recordings or reproduced albums. This paper studies two-dimensional Fourier transform as a feature-dimension reduction method to search cover song fast. The two-dimensional Fourier transform is conducive in feature-dimension reduction for cover song search due to musical-key invariance. This paper extends the previous work, which only utilize the magnitude of the Fourier transform, by introducing an invariant from phase based on the assumption that adjacent frames have the same musical-key change. We compare the cover song retrieval accuracy of the Fourier-transform based methods over two datasets. The experimental results show that the addition of the invariant from phase improves the cover song retrieval accuracy over the previous magnitude-only method.

• Journal of Korea Multimedia Society
• /
• v.22 no.6
• /
• pp.639-645
• /
• 2019

This paper proposes a salient chromagram components extraction method based on the temporal discrete cosine transform of a chromagram block to improve cover song retrieval accuracy. The proposed salient chromagram emphasizes tonal contents of music, which are well-preserved between an original song and its cover version, while reducing the effects of timbre difference. We apply the proposed salient chromagram extraction method as a preprocessing step for the Fourier-transform based cover song matching. Experiments on two cover song datasets confirm that the proposed salient chromagram improves the cover song search accuracy.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.11a
• /
• pp.850-853
• /
• 2020

gradient decent 를 기반으로 한 Differentiable architecture search(DARTS)는 한 번의 Architecture Search 로 모든 후보 연산 중 가장 가중치가 높은 연산 하나를 선택한다. 이 때 비슷한 종류의 연산이 가중치를 나누어 갖는 "표의 분산"이 나타나, 성능이 더 좋은 연산이 선택되지 못하는 상황이 발생한다. 본 연구에서는 이러한 상황을 막기위해 Architecture Parameter 가중치의 gradient 를 기반으로 연산들을 클러스터링 하여 그룹화 한다. 그 후 그룹별로 가중치를 합산하여 높은 가중치를 갖는 그룹만을 사용하여 한 번 더 Architecture Search 를 진행한다. 각각의 Architecture Search 는 DARTS 의 절반 epoch 만큼 이루어지며, 총 epoch 이 같으나 두번째의 Architecture Search 는 선별된 연산 그룹을 사용하므로 DARTS 에 비해 더 적은 Search Cost 가 요구된다. "표의 분산"문제를 해결하고, 2 번으로 나뉜 Architecture Search 에 따라 CIFAR 10 데이터 셋에 대해 2.46%의 에러와 0.16 GPU-days 의 탐색시간을 얻을 수 있다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.8
• /
• pp.3516-3541
• /
• 2018

Task scheduling is one of the most challenging aspects of cloud computing nowadays, and it plays an important role in improving overall performance in, and services from, the cloud, such as response time, cost, makespan, and throughput. A recent cloud task-scheduling algorithm based on the symbiotic organisms search (SOS) algorithm not only has fewer specific parameters, but also incurs time complexity. SOS is a newly developed metaheuristic optimization technique for solving numerical optimization problems. In this paper, the basic SOS algorithm is reduced, and chaotic local search (CLS) is integrated into the reduced SOS to improve the convergence rate. Simulated annealing (SA) is also added to help the SOS algorithm avoid being trapped in a local minimum. The performance of the proposed SA-CLS-SOS algorithm is evaluated by extensive simulation using the Matlab framework, and is compared with SOS, SA-SOS, and CLS-SOS algorithms. Simulation results show that the improved hybrid SOS performs better than SOS, SA-SOS, and CLS-SOS in terms of convergence speed and makespan.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.6 no.1
• /
• pp.91-103
• /
• 2010

This paper introduces efficient rule-based semantic search techniques to ontology-based knowledge commerce services. Primarily, the search techniques presented in this paper define rules of reasoning that are required for users to search using the concept of ontology, multiple characteristics, relations among concepts and data type. In addition, based on the defined rules, the rule-based reasoning techniques search ontology for knowledge commerce services. This paper explains the conversion rules of query which convert user's query language into semantic search words, and transitivity rules which enable users to search related tags, knowledge products and users. Rule-based sematic search techniques are also presented; these techniques comprise knowledge search modules that search ontology using validity examination of queries, query conversion modules for standardization and expansion of search words and rule-based reasoning. The techniques described in this paper can be applied to sematic knowledge search systems using tags, since transitivity reasoning, which uses tags, knowledge products, and relations among people, is possible. In addition, as related users can be searched using related tags, the techniques can also be employed to establish collaboration models or semantic communities.

• Korean Journal of Health Education and Promotion
• /
• v.30 no.3
• /
• pp.59-73
• /
• 2013

Objectives: The study is aimed at examining the individual reasons and regional/environmental factors of online search on suicide using social big data to predict practical behaviors related to suicide and to develop an online suicide prevention system on the governmental level. Methods: The study was conducted using suicide-related social big data collected from online news sites, blogs, caf$\acute{e}$s, social network services and message boards between January 1 and December 31, 2011 (321,506 buzzes from users assumed as adults and 67,742 buzzes from those assumed as teenagers). Technical analysis and development of the suicide search prediction model were done using SPSS 20.0, and the structural model, nd multi-group analysis was made using AMOS 20.0. Also, HLM 7.0 was applied for the multilevel model analysis of the determinants of search on suicide by teenagers. Results: A summary of the results of multivariate analysis is as follows. First, search on suicide by adults appeared to increase on days when there were higher number of suicide incidents, higher number of search on drinking, higher divorce rate, lower birth rate and higher average humidity. Second, search on suicide by teenagers rose on days when there were higher number of teenage suicide incidents, higher number of search on stress or drinking and less fine dust particles. Third, the comparison of the results of the structural equation model analysis of search on suicide by adults and teenagers showed that teenagers were more likely to proceed from search on stress to search on sports, drinking and suicide, while adults significantly tended to move from search on drinking to search on suicide. Fourth, the result of the multilevel model analysis of determinants of search on suicide by teenagers showed that monthly teenagers suicide rate and average humidity had positive effect on the amount of search on suicide. Conclusions: The study shows that both adults and teenagers are influenced by various reasons to experience stress and search on suicide on the Internet. Therefore, we need to develop diverse school-level programs that can help relieve teenagers of stress and workplace-level programs to get rid of the work-related stress of adults.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2016.01a
• /
• pp.33-35
• /
• 2016

본 논문에서는 모바일 디바이스를 이용하여 촬영한 이미지 또는 이미 모바일 디바이스에 저장된 이미지를 사용자가 검색을 위한 질의어로 사용할 수 있는 방법에 대하여 제안한다. 기존의 모바일 검색엔진을 그대로 활용하기 위해 이미지 어노테이션에 기반한 태깅 키워드를 검색 이미지와 매칭하여 질의하는 방식으로 구현하며, 이 과정에서 이미지의 분석과 분류를 위한 SVM(Support Vector Machine)과 SIFT(Scale Invariant Feature Transform) 알고리즘을 사용하였으며, 이미지 어노테이션 태깅에 대한 키워드 매칭을 위해 빅데이터에서의 MapReduce를 응용하였다.

• Korean Journal of Oriental Medicine
• /
• v.17 no.2
• /
• pp.57-62
• /
• 2011

We in this paper propose a semantic search system using the basic ontology in Korean medicine field. The basic ontology provides a formalization of medicinal materials, formulas, and diseases of Korean medicine. Recently, many studies for the semantic search system have been proposed. However, they do not support the semantic search and reasoning in the domain of Korean medicine because they do not have the Korean medicine ontology. Our system provides the semantic search features of semantic keyword recommendation, associated information browsing, and ontology reasoning based on the basic ontology. In addition, they also have the features of ontology search of a form of table and graph, synonym search, and external Open API supports. The general search engines usually provide search results for the simple keyword, while our system can also provide the associated information with respect to search results by using ontology so that can recommend more exact results to users.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.5
• /
• pp.195-200
• /
• 2021

Due to the development of video sharing platforms, the amount of video uploads is exploding. Such videos often include various types of music, among which cover songs are included. In order to protect the copyright of music, an algorithm to find the original song of the cover song is essential. However, it is not easy to find the original song because the cover song is a modification of the composition, speed and overall structure of the original song. So far, there is no known effective algorithm for searching the original song of the cover song. In this paper, we propose an algorithm for searching the original song of the cover song using the inflection points of the melody line. Inflection points represent the characteristic points of change in the melody sequence. The proposed algorithm compares the original song and the cover song using the sequence of inflection points for the representative phrase of the original song. Since the characteristics of the representative phrase are used, even if the cover song is a song made by modifying the overall composition of the song, the algorithm's search performance is excellent. Also, since the proposed algorithm uses only the features of the inflection point sequence, the memory usage is very low. The efficiency of the algorithm was verified through performance evaluation.