• Proceedings of the Korean Society for Information Management Conference
• /
• 1994.12a
• /
• pp.15-18
• /
• 1994

본 연구에서는 불논리검색의 단점을 보완하기 위한 가장 강력한 검색 모형인 퍼지검색과 확률검색의 효율을 불논리검색과 상호비교하였다. 실험데이터로 정보학 분야의 한국어 test collection인 KT Test Set을 이용하였고 색인어와 색인어의 문헌내 출현빈도를 바탕으로 퍼지시소러스를 생성하여 시소러스의 NT, BT로 탐색식을 확장한 다음 각각에 대해 3가지 검색을 행하고 검색효율을 평균재현율과 평균정확률로 측정하였다. 실험결과 검색효율은 재현율에서는 확률검색, 불논리검색, 퍼지검색 순으로. 정확률에서는 퍼지검색, 확률검색, 불논리검색 순으로 나타났다.

• Journal of the Korean Society for information Management
• /
• v.16 no.3
• /
• pp.49-63
• /
• 1999

The purpose of this study is to investigate whether the terms in queries are distributed according to the 2-Poisson model in the documents represented by abstract/title or full-text. In this study, retrieval experiments using Binary independence and 2-Poisson independence model, which are based on the probabilistic theory, were conducted to see if the 2-Poisson distribution of the query terms has an influence on the retrieval effectiveness, particularly of full-text information retrieval system.

• The Korean Journal of Applied Statistics
• /
• v.32 no.5
• /
• pp.763-782
• /
• 2019

This paper proposes a probabilistic document ranking model incorporating term dependencies. Document ranking is a fundamental information retrieval task. The task is to sort documents in a collection according to the relevance to the user query (Qin et al., Information Retrieval Journal, 13, 346-374, 2010). A probabilistic model is a model for computing the conditional probability of the relevance of each document given query. Most of the widely used models assume the term independence because it is challenging to compute the joint probabilities of multiple terms. Words in natural language texts are obviously highly correlated. In this paper, we assume a multinomial distribution model to calculate the relevance probability of a document by considering the dependency structure of words, and propose an information retrieval model to rank a document by estimating the probability with the maximum entropy method. The results of the ranking simulation experiment in various multinomial situations show better retrieval results than a model that assumes the independence of words. The results of document ranking experiments using real-world datasets LETOR OHSUMED also show better retrieval results.

• Journal of the Korea Computer Industry Society
• /
• v.2 no.2
• /
• pp.203-210
• /
• 2001

To estimate relational models and test the theoretical hypotheses of binary tree search algorithms, we built binary search trees with random permutations of n (number of nodes) distinct numbers, which ranged from three to seven. Probabilities for building binary search trees corresponding to each possible height and balance factor were estimated. Regression models with variables of number of nodes, height, and average number of comparisons were estimated and the theorem of O(1g(n)) was accepted experimentally by a Lack of Test procedure. Analysis of Variance model was applied to compare the average number of comparisons with three groups by height and balance factor of the trees to test theoretical hypotheses of a binary search tree performance statistically.

• The Korean Journal of Applied Statistics
• /
• v.13 no.1
• /
• pp.1-10
• /
• 2000

정보화사회에서 효율적인 정보검색(information retrieval)은 각종 의사결정에 매우 중요하다. 주어진 정보검색 문제가 있을 때 과거에 검색되었던 자료는 그 적절성 여부에 대한 평가를 데이터베이스에 첨가하여 지식베이스(knowledge base)화 할 수 있다. 본 연구는 이 지식베이스에 대한 논리적 패턴을 분석하여 새로운 정보의 '적정성(relevance)' 여부를 판별하는 확률화 정보검색 모형을 만들고 이에 대한 실험을 하였다.

• Journal of the Korean Society for information Management
• /
• v.10 no.2
• /
• pp.43-67
• /
• 1993

The primary purpose of the study is to improve the low preclslon often found In full-text retrleval systems. In order to enhance the low precision of full-text retrleval wh~le retaining ~ t s hgh recall, relevance feedback mechanisms based on probabilistic retrieval models (binary independence and two-Polsson Independence models) were employed. Thls paper investigates the effect of relevance feedback on the performance of full-text retrieval systems.

• Journal of the Korea Computer Industry Society
• /
• v.3 no.2
• /
• pp.235-242
• /
• 2002

To minimize restructuring cost of a tree, experiments were conducted to collect quantitative information of searching efficiencies varying degrees of imbalance in a binary search tree. Degrees of tree imbalance were measured by a balance factor, an absolute value of height difference of left subtree and right subtree in a binary search tree. The average number of comparisons increased (p＜0.01), and searching efficiency of O(n) was more appropriate rather than O(logn), as degrees of imbalance in a binary search tree deteriorated. However, there were no significant differences of searching efficiencies in height balanced trees and trees with subtrees to have height 3 less than the other (p＞0.05). Therefore, the findings would be applicable to maintain searching efficiency of a software with a binary search tree.

• Proceedings of the Korean Society for Information Management Conference
• /
• 1999.08a
• /
• pp.113-116
• /
• 1999

인터넷 이용 경험이 적은 초보자뿐만 아니라 경험자도 인터넷에서 자신에게 필요한 정보를 찾기는 쉽지 않다. 인터넷 정보자원은 급증하는 반면 인터넷 검색엔진이 원하는 정보를 찾아주는 확률은 오히려 줄어들고 있는데 이는 체계적으로 정리되어 있지 않기 때문이다. 본 연구는 인터넷의 잔문 즉 웹브라우저를 실행한 후 처음 접속하는 사이트인 포털사이트(Portal site)를 전문화하여 궁극적으로 포털을 정보이용의 최종 목적지가 될 수 있는 모형을 설계하였다. 이 모형은 도서관이 보유하고 있는 학술정보 뿐만 아니라 인터넷의 정보자원을 체계적으로 정리하고 부가서비스를 제공하여 이용자의 정보이용 편의성을 높이는데 있다.

• Journal of Digital Convergence
• /
• v.11 no.6
• /
• pp.175-185
• /
• 2013

For effective information sharing, data provenance semantics need to be managed effectively. Based on a scheme to represent data provenance semantics, we propose a model to calculate information sharing costs. Information sharing costs are derived from probabilities of type I and type II errors that occur in organizational information sharing, costs related to these errors, and information sharing distances between organizational units which are determined by information sharing workflows. We apply the model to various types of information sharing workflows including departmental information systems, hierarchical information systems, a hub and a stand-alone system. The calculated information sharing costs show that the hub with data standardization is best in information sharing; however without standardization its information sharing cost deteriorates to that of a departmental information system. And, any information sharing workflow is better than a stand-alone system. It is proved that the model is useful in analyzing effectiveness of information sharing workflows and their characteristics.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.6
• /
• pp.1279-1286
• /
• 2021

This study suggests the user's needs for a smart parking service that enables parking lot search and advance reservation service, and is a study on the user's preference selection model related to fees (reservation fee, penalty fee), etc. Two types of user preference models in the form of logit models were constructed by composing a response questionnaire for smart parking service. The first is a model for selecting a smart parking lot, which suggests a situation in which the probability of selection is higher than that of a general parking lot in the relationship between usage fee and cost. The second is a parking ticket reservation discount selection model, and the smart parking service selection probability was analyzed through the relationship model between the reservation amount and the penalty. It can be used as a design requirement that enables sophisticated and various types of smart parking service considering users' preferences.