• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.171-176
• /
• 2023

최근 등장한 대규모 데이터로 사전학습된 자연어 생성 모델들은 대화 능력 및 코드 생성 태스크등에서 인상적인 성능을 보여주고 있어, 본 논문에서는 대형 언어 모델 (LLM)의 한국어 질문을 SQL 쿼리 (Text-to-SQL) 변환하는 성능을 평가하고자 한다. 먼저, 영어 Text-to-SQL 벤치마크 데이터셋을 활용하여 영어 질의문을 한국어 질의문으로 번역하여 한국어 Text-to-SQL 데이터셋으로 만들었다. 대형 생성형 모델 (GPT-3 davinci, GPT-3 turbo) 의 few-shot 세팅에서 성능 평가를 진행하며, fine-tuning 없이도 대형 언어 모델들의 경쟁력있는 한국어 Text-to-SQL 변환 성능을 확인한다. 또한, 에러 분석을 수행하여 한국어 문장을 데이터베이스 쿼리문으로 변환하는 과정에서 발생하는 다양한 문제와 프롬프트 기법을 활용한 가능한 해결책을 제시한다.

• Electronics and Telecommunications Trends
• /
• v.11 no.1 s.39
• /
• pp.1-8
• /
• 1996

웹 서비스가 인터넷을 주도하면서 Hyper Text Markup Language(HTML) 문서 위주의 정보검색 서비스에서 한 발 더 나아가 SQL 데이터베이스와 웹을 통합하여 이를 비즈니스 분야로 확대 발전시키기 위한 연구가 진행되고 있다. 실제로 많은 웹 서버에서 Common Gateway Interface(CGI) 기능을 데이터베이스 검색에 활용하고 있다. 이와 함께 GCI를 사용자 입장에서 한 단계 발전시킨 WWW interface to DataBase(WDB), Gateway Structure Query Language(GSQL) 등이 인터넷에서 소개되고 있다. 본 고에서는 현재 다각도로 진행중인 웹과 데이터베이스 통합기술 동향을 CGI와 WDB를 통해 살펴보았다.

• Proceedings of the Korea Contents Association Conference
• /
• 2004.11a
• /
• pp.442-447
• /
• 2004

XML is normalized text form that is designed to transmit structured document in web as that propose in W3C (World Wide Web Consortium) in 1996. Function that this can overcome HTML's limit that use in existing in Internet and user define new tag to HTML by way to solve SGML's complexity added. There is many efforts to use storing this XML document in RDBMS but to relation style DB because XML document is tree structure structurally data SQL and perfect disaster caused by things that is language to ask a question accomplish XQuery that so it is W3C's XML standard query appear. After store XML informations including space information to RDBMS in this paper, Spatial XQuery through converter that is Sqatial XQuery2SQL through Spatial operator, Spatial function SQL of by Sqatial XQuery2SQL conversion algorithm that draw information in RDBMS after change embody wish to.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.3
• /
• pp.430-433
• /
• 2007

There have been several researches to manipulate XML Queries efficiently since XML has been accepted in many areas. Among the many of the researches majority of them adopt relational databases as underlying systems because relational model which is used the most widely for managing large data efficiently. In this paper we develop XQuery to SQL Translation Algorithms called XSTAR that can efficiently handle XPath, XQuery FLWORs with nested iteration expressions, element constructors and keywords retrieval on relational database as well as constructing XML fragments from the transformed SQL results. The entire algorithms mentioned in XSTAR have been implemented as the XQuery processor engine in XML management system, XPERT, and we can test and confirm it's prototype from "http ://dblab.kmu.ac.kr/project.jsp".

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09b
• /
• pp.49-52
• /
• 2003

In this research, we propose a mechanism to develop an inference engine and expert systems based on relational database and SQL (structured query language). Generally, former researchers had tried to develop an expert systems based on text-oriented knowledge base and backward/forward (chaining) inference engine. In these researches, however, the speed of inference was remained as a tackling point in the development of agile expert systems. Especially, the forward inference needs more times than backward inference. In addition, the size of knowledge base, complicate knowledge expression method, expansibility of knowledge base, and hierarchies among rules are the critical limitations to develop an expert systems. To overcome the limitations in speed of inference and expansibility of knowledge base, we proposed a relational database-oriented knowledge base and forward inference engine. Therefore, our proposed mechanism could manipulate the huge size of knowledge base efficiently, and inference with the large scaled knowledge base in a short time. To this purpose, we designed and developed an SQL-based forward inference engine using relational database. In the implementation process, we also developed a prototype expert system and presented a real-world validation data set collected from medical diagnosis field.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.6
• /
• pp.743-748
• /
• 2003

In this research, we propose a mechanism to develop an inference engine and expert systems based on relational database (RDB) and SQL (structured query language). Generally, former researchers had tried to develop an expert systems based on text-oriented knowledge base and backward/forward (chaining) inference engine. In these researches, however, the speed of inference was remained as a tackling point in the development of agile expert systems. Especially, the forward inference needs more times than backward inference. In addition, the size of knowledge base, complicate knowledge expression method, expansibility of knowledge base, and hierarchies among rules are the critical limitations to develop an expert system. To overcome the limitations in speed of inference and expansibility of knowledge base, we proposed a relational database-oriented knowledge base and forward inference engine. Therefore, our proposed mechanism could manipulate the huge size of knowledge base efficiently. and inference with the large scaled knowledge base in a short time. To this purpose, we designed and developed an SQL-based forward inference engine using relational database. In the implementation process, we also developed a prototype expert system and presented a real-world validation data set collected from medical diagnosis field.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.7 no.4
• /
• pp.274-278
• /
• 2007

As XML becomes the standard of digital data exchange format we need to manage a lot of XML data effectively. Unlike tables in relational model XML documents are not structural. That makes it difficult to store XML documents as tables in relational model. To solve these problems there have been significant researches in relational database systems. There are two kinds of approaches: 1) One way is to decompose XML documents so that elements of XML match fields of relational tables. 2) The other one stores a whole XML document as a field of relational table. In this paper we adopted the second approach to store XML documents because sometimes it is not easy for us to decompose XML documents and in some cases their element order in documents are very meaningful. We suggest an efficient table schema to store only inverted index as tables to retrieve required data from XML data fields of relational tables and shows SQL translations that correspond to XML full-text retrievals. The functionalities of XML retrieval are based on the W3C XQuery which includes full-text retrievals. In this paper we show the superiority of our method by comparing the performances in terms of a response time and a space to store inverted index. Experiments show our approach uses less space and shows faster response times.

• Journal of Korea Multimedia Society
• /
• v.8 no.2
• /
• pp.154-163
• /
• 2005

Because most researches that were studied so far on XML documents used an absolute coordinate system in most of the index techniques, the update operation makes a large burden. To express the structural relations between elements, attributes and text, we need to reconstruct the structure of the coordinates. As the reconstruction process proceeds through out the entire XML document in a cascade manner, which is not limited to the current changing node, a serious performance problem may be caused by the frequent update operations. In this paper, we propose an index technique based on extensible index that does not cause serious performance degradations. It can limit the number of node to participate in reconstruction process and improve lots of performance capacities on the whole. And extensible index performs the containment relationship query by the simple expression using SQL statement.

• Journal of Korea Multimedia Society
• /
• v.11 no.1
• /
• pp.1-12
• /
• 2008

Database encryption is one of the important mechanisms for prohibiting internal malicious users and outside hackers from utilizing data. Frequent occurrences of encryption and decryption cause degradation of database performance so that many factors should be considered in implementing encryption system. In this paper, we propose an architecture of database encryption system and data encryption module. In addition we suggest extended SQL in order to manage data encryption and decryption. In implementing database encryption system, we adopt ARIA encryption algorithm which is proved to be the most fast one among Korea standardized encryption algorithm. We use an single key for each database in encrypting data rather than using several keys in order to improve performance. Research over performance evaluation of database encryption system is rare up to now. Based on our implemented system, we provide performance evaluation results over various H/W platforms and compare performance differences between plain text and encrypted data.

• Journal of information and communication convergence engineering
• /
• v.7 no.3
• /
• pp.275-280
• /
• 2009

While database(DB) and information retrieval(IR) have been developed independently, there have been emerging requirements that both data management and efficient text retrieval should be supported simultaneously in an information system such as health care, customer support, XML data management, and digital libraries. The great divide between DB and IR has caused different manners in index maintenance for newly arriving documents. While DB has extended its SQL layer to cope with text fields due to lack of intact mechanism to build IR-like index, IR usually treats a block of new documents as a logical unit of index maintenance since it has no concept of integrity constraint. However, In the DB-IR integrations, a transaction on adding or updating a document should include maintenance of the posting lists accompanied by the document. Although DB-IR integration has been budded in the research filed, the issue will remain difficult and rewarding areas for a while. One of the primary reasons is lack of efficient online transactional index maintenance. In this paper, performance of a few strategies for per-document basis transactional index maintenance - direct index update, pulsing auxiliary index and posting segmentation index - will be evaluated. The result shows that the pulsing auxiliary strategy and posting segmentation indexing scheme, can be a challenging candidates for text field indexing in DB-IR integration.