• Journal of Industrial Technology
• /
• v.28 no.B
• /
• pp.209-214
• /
• 2008

Set-valued attributes appear in many applications to model complex objects occurring in the real world. One of the most important operations on set-valued attributes is the set join, because it provides a various method to express complex queries. Currently proposed set join algorithms are based on block nested loop join in which inverted files are partitioned horizontally into blocks. Evaluating these joins are expensive because they generate intermediate partial results severely and finally obtain the final results after merging partial results. In this paper, we present an efficient processing of set join algorithm. We propose a new set join algorithm that vertically partitions inverted files into blocks, where each block fits in memory, and performs block nested loop join without producing intermediate results. Our experiments show that the vertical bitmap nested set join algorithm outperforms previously proposed set join algorithms.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.7
• /
• pp.3272-3285
• /
• 2016

Recently, Vector of Locally Aggregated Descriptors (VLAD) has been proposed to index image by compact representations, which encodes powerful local descriptors and makes significant improvement on search performance with less memory compared against the state of art. However, its performance relies heavily on the size of the codebook which is used to generate VLAD representation. It indicates better accuracy needs higher dimensional representation. Thus, more memory overhead is needed. In this paper, we enhance VLAD image representation by using two level hierarchical-codebooks. It can provide more accurate search performance while keeping the VLAD size unchanged. In addition, hierarchical-codebooks are used to construct multiple inverted files for more accurate non-exhaustive search. Experimental results show that our method can make significant improvement on both VLAD image representation and non-exhaustive search.

• The KIPS Transactions:PartA
• /
• v.17A no.5
• /
• pp.229-236
• /
• 2010

With an increasing number of documents in the Internet or enterprises, it becomes crucial to efficiently support users' queries on those documents. In that situation, the full-text search technique is accepted in general, because it can answer uncontrolled ad-hoc queries by automatically indexing all the keywords found in the documents. The size of index files made for full-text searches grows with the increasing number of indexed documents, and thus the disk cost may be too large to process multi-keyword queries against those enlarged index files. To solve the problem, we propose both of the index file structure and its management scheme suitable to the processing of multi-keyword queries against a large volume of index files. For this, we adopt the structure of inverted-files, which are widely used in the multi-keyword searches, as a basic index structure and modify it to a hierarchical structure for join operations and ranking operations performed during the query processing. In order to save disk costs based on that index structure, we dynamically store in the main memory the results of join operations between two keywords, if they are highly expected to be entered in users' queries. We also do performance comparisons using a cost model of the disk to show the performance advantage of the proposed scheme.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.10c
• /
• pp.35-40
• /
• 2007

역파일을 사용한 색인 기법은 정보 검색 분야에서 널리 사용되었으며, 최근 대용량 검색 시스템으로 사용되고 있는 웹 검색 엔진에서도 적응되고 있다. 본 논문에서는 웹 검색 엔진의 특성에 완친 구현된 역파일 기법 기반의 웹 문서 색인 파일의 구조와 디스크에 저장된 대용량의 역파일 색인을 기반으로 웹 페이지의 검색 적합도를 계산하는 랭킹 시스템을 설명한다. 이를 통하여 상용 웹 검색 엔진의 랭킹 시스템과 디스크 자원 사용의 최소화 기법을 제시한다.

• Journal of the Korean Society for information Management
• /
• v.18 no.3
• /
• pp.29-40
• /
• 2001

In this paper, we suggest a query optimization algorithm to select the optimal processing method of an extended boolean query on inverted files. There can be a lot of methods for processing an extended boolean query according to the processing sequence oh the keywords con tamed in the query, In this sense, the problem of optimizing an extended boolean query it essentially that of optimizing the keyword sequence in the query. In this paper, we show that the problem is basically analogous to the problem of finding the optimal join order in database query optimization, and apply the ideas in the area to the problem solving. We establish the cost model for processing an extended boolean query and develop an algorithm to filled the optimal keyword-processing sequence based on the concept of keyword rank using the keyword selectivity and the access costs of inverted file. We prove that the method selected by the optimization algorithm is really optimum, and show, through experiments, that the optimal method is superior to the others in performance We believe that the suggested optimization algorithm will contribute to the significant enhancement of the information retrieval performance.

• Journal of Internet Computing and Services
• /
• v.11 no.3
• /
• pp.75-86
• /
• 2010

As the interests in health grow higher, necessity of Well-being relation informations get more importance. We get the information of well-being, tinternet retrieval system or blog, homepage and media. Although, it is not easy to find informations of well-being food. So, retrieval system has been requiring information about well-being food. In this paper, Weight-based Wellbeing Food Retrieval System is designed and implemention. Finding numerous pages and if well-being keywords includes page, it was identified and add weight. User searching for keywords, it implement, well-being food pages comes at the first. Keywords for discrimination makes type of dictionary, so it can insert, delete, modify. Inverted files saves hasing(direct-based file). Retrieval System in this paper is experimental result, at keywords of well-being food show 5~15% imprement than another Retrieval System. In this paper, Weight-based Wellbeing Food Retrieval System's designed and proposed way to raking for well-being food.

• Journal of the Korean Society for information Management
• /
• v.21 no.3
• /
• pp.1-14
• /
• 2004

Visualization of documents will help users when they do search similar documents. and all research in information retrieval addresses itself to the problem of a user with an information need facing a data source containing an acceptable solution to that need. In various contexts. adequate solutions to this problem have included alphabetized cubbyholes housing papyrus rolls. microfilm registers. card catalogs and inverted files coded onto discs. Many information retrieval systems rely on the use of a document surrogate. Though they might be surprise to discover it. nearly every information seeker uses an array of document surrogates. Summaries. tables of contents. abstracts. reviews, and MARC recordsthese are all document surrogates. That is, they stand infor a document allowing a user to make some decision regarding it. whether to retrieve a book from the stacks, whether to read an entire article, etc. In this paper another type of document surrogate is investigated using a grouping method of term list. lising Multidimensional Scaling Method (MDS) those surrogates are visualized on two-dimensional graph. The distances between dots on the two-dimensional graph can be represented as the similarity of the documents. More close the distance. more similar the documents.