• Journal of Korea Multimedia Society
• /
• v.22 no.9
• /
• pp.1000-1010
• /
• 2019

Image semantic segmentation and dehazing are key tasks in the computer vision. In recent years, researches in both tasks have achieved substantial improvements in performance with the development of Convolutional Neural Network (CNN). However, most of the previous works for semantic segmentation assume the images are captured in clear weather and show degraded performance under hazy images with low contrast and faded color. Meanwhile, dehazing aims to recover clear image given observed hazy image, which is an ill-posed problem and can be alleviated with additional information about the image. In this work, we propose a deep multi-task network for simultaneous semantic segmentation and dehazing. The proposed network takes single haze image as input and predicts dense semantic segmentation map and clear image. The visual information getting refined during the dehazing process can help the recognition task of semantic segmentation. On the other hand, semantic features obtained during the semantic segmentation process can provide cues for color priors for objects, which can help dehazing process. Experimental results demonstrate the effectiveness of the proposed multi-task approach, showing improved performance compared to the separate networks.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.5
• /
• pp.1807-1824
• /
• 2015

As the technologies advance, mobile peer-to-peer (MP2P) networks or systems become one of the major ways to share resources and information. On such a system, the information retrieval (IR), including the development of scalable infrastructures for indexing, becomes more complicated due to a huge increase on the amount of information and rapid information change. To keep the systems on MP2P networks more reliable and consistent, the index structures need to be updated frequently. For a semantic IR system, the index structure is even more complicated than a classic IR system and generally has higher update cost. The most well-known indexing technique used in semantic IR systems is Latent Semantic Indexing (LSI), of which the index structure is generated by singular value decomposition (SVD). Although LSI performs well, updating the index structure is not easy and time consuming. In an MP2P environment, which is fully distributed and dynamic, the update becomes more challenging. In this work, we consider how to update the sematic index generated by LSI and keep the index consistent in the whole MP2P network. The proposed Concept Space Update (CSU) protocol, based on distributed 2-Phase locking strategy, can effectively achieve the objectives in terms of two measurements: coverage speed and update cost. Using the proposed effective synchronization mechanism with the efficient updates on the SVD, re-computing the whole index on the P2P overlay can be avoided and the consistency can be achieved. Simulated experiments are also performed to validate our analysis on the proposed CSU protocol. The experimental results indicate that CSU is effective on updating the concept space with LSI/SVD index structure in MP2P semantic IR systems.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.04b
• /
• pp.238-240
• /
• 2002

If there is an Information Retrieval system which comprehends the semantic content of documents and knows the preference of users. the system can search the information better on the Internet, or improve the IR performance. Therefore we propose the IR model which combines semantic based indexing and fuzzy relevance model. In addition to the statistical approach, we chose the semantic approach in indexing, lexical chains, because we assume it would improve the performance of the index term extraction. Furthermore, we combined the semantic based indexing with the fuzzy model, which finds out the exact relevance of the user preference and index terms. The proposed system works as follows: First, the presented system indexes documents by the efficient index term extraction method using lexical chains. And then, if a user tends to retrieve the information from the indexed document collection, the extended IR model calculates and ranks the relevance of user query. user preference and index terms by some metrics. When we experimented each module, semantic based indexing and extended fuzzy model. it gave noticeable results. The combination of these modules is expected to improve the information retrieval performance.

• The KIPS Transactions:PartA
• /
• v.18A no.3
• /
• pp.109-114
• /
• 2011

We describe a technique for automatically proving compiler optimizations sound, meaning that their transformations are always semantics-preserving. As is well known, IR (Intermediate Representation) optimization is an important step in a compiler backend. But unfortunately, it is difficult to detect and debug the IR optimization errors for compiler developers. So, we introduce a C level error check system for detecting the correctness of these IR transformation techniques. In our system, we first create an IR-to-C converter to translate IR to C code before and after each compiler optimization phase, respectively, since our technique is based on the Memory Comparison-based Clone(MeCC) detector which is a tool of detecting semantic equivalency in C level. MeCC accepts only C codes as its input and it uses a path-sensitive semantic-based static analyzer to estimate the memory states at exit point of each procedure, and compares memory states to determine whether the procedures are equal or not. But MeCC cannot guarantee two semantic-equivalency codes always have 100% similarity or two codes with different semantics does not get the result of 100% similarity. To increase the reliability of the results, we describe a technique which comprises how to generate C codes in IR-to-C transformation phase and how to send the optimization information to MeCC to avoid the occurrence of these unexpected problems. Our methodology is illustrated by three familiar optimizations, dead code elimination, instruction scheduling and common sub-expression elimination and our experimental results show that the C level error check system is highly reliable.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.7
• /
• pp.19-29
• /
• 2019

Requirements traceability is regarded as one of the important quality attributes in software requirements engineering field. If requirements traceability is guaranteed then we can trace the requirements' life throughout all the phases, from the customers' needs in the early stage of the project to requirements specification, deployment, and maintenance phase. This includes not only tracking the development artifacts that accompany the requirements, but also tracking backwards from the development artifacts to the initial customer requirements associated with them. In this paper, especially, we dealt with the traceability between functional requirements and non-functional requirements. Among many Information Retrieval (IR) techniques, we decided to utilize Latent Semantic Indexing (LSI) and Latent Dirichlet Allocation (LDA) in our research. Ultimately, we conducted an experiment on constructing traceability by using two techniques and analyzed the experiment results. And then we provided a comparative study between two IR techniques in constructing traceability between functional requirements and non-functional requirements.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.23 no.1
• /
• pp.91-111
• /
• 2012

With the rapid expansion of web information, conventional information retrieval techniques are becoming inadequate for users and often result in disappointment, because a couple of simple keywords can easily produce information too much. This study aims at the development of Web information retrieval techniques based on semantics to improve the quality of understanding for information. To achieve the goal, this study analyzes technologies and current status of researches on semantic information retrieval. With the results which are requirements, system architecture and indexing method, this study proposes the system architecture of semantic-based information retrieval system.

• Language and Information
• /
• v.6 no.2
• /
• pp.83-104
• /
• 2002

I argue that Definiteness Effect (DE) should include Indefiniteness Restriction (IR) as well as Definiteness Restriction (DR). DR is exhibited by existential constructions, predicate nominals, inalienable possession constructions, and verbs with semantic features like 〔((CAUSE TO) COME TO) EXIST〕. IR is caused by some existence presupposition of aspectual adverbs, aspectual verbs, repetitives, and topic markers. The environments for DR and IR determine the (in)definiteness of Korean bare noun phrases which otherwise can be used either way. The neutralization of DR is also induced by focus which imposes a certain amount of structure on the event quantification. Van der Sandt's (1992) Presuppositions-as- Anaphora-Theory is effectively used to account for those examples from DR, ID, and the neutralization of BR.

• Journal of KIISE:Software and Applications
• /
• v.29 no.8
• /
• pp.584-594
• /
• 2002

In information retrieval or digital library, one of the most important factors is to find out the exact information which users need. In this paper, we present an efficient index term extraction method which makes it possible to guess the content of documents and get the information more exactly. To find out index terms in a document, we use lexical chains. Before generating lexical chains, we roughly disambiguate the senses of nouns in a document using specific concept, called semantic window. Semantic window is that we look ahead semantic relations of peripheral nouns and disambiguate the senses of nouns. After generating lexical chains with sense-disambiguated nouns, we find out strong chains by some metrics and extract index terms from a few strong chains. We evaluated our system, using results of a key phrase extraction system, KEA. This system works in general domains of documents Including Information Retrieval and Digital Library.

• Journal of KIISE:Databases
• /
• v.34 no.6
• /
• pp.554-563
• /
• 2007

The e-Marketplace is growing rapidly and providing a more complex relationship between providers and consumers. In recent years, e-Marketplace integration or cooperation issues have become an important issue in e-Business. The e-Catalog is a key factor in e-Business, which means an e-Catalog System needs to contain more large data and requires a more efficient retrieval system. This paper focuses on designing an efficient retrieval system for very large e-Catalogs of large e-Marketplaces. For this reason, a new similarity measure for e-Catalog retrieval based on semantic relationships was proposed. Our achievement is this: first, a new e-Catalog data model based on semantic relationships was designed. Second, the model was extended by considering lexical features (Especially, focus on Korean). Third, the factors affecting similarity with the model was defined. Fourth, from the factors, we finally defined a new similarity measure, realized the system and verified it through experimentation.

• Journal of Korea Society of Industrial Information Systems
• /
• v.12 no.4
• /
• pp.25-31
• /
• 2007

An automatic term-network construction system is proposed in this paper. This system uses the statistical values of the terms appeared in a document corpus. The 186 oral history documents collected from the Saemangeum area of Chollapuk-do, Korea, are used for the research. The term relationships presented in the term-network are decided by the cosine similarities of the term vectors. The number of the terms extracted from the documents is about 1700. The system is able to show the term relationships from the term-network as quickly as like a real-time system. The way of this term-network construction is expected as one of the methods to construct the ontology system and to support the semantic retrieval system in the near future.