• Journal of Information Science Theory and Practice
• /
• 제3권1호
• /
• pp.24-39
• /
• 2015

Pre-retrieval query formulation is an important step for identifying local text reuse. Local reuse with high obfuscation, paraphrasing, and translation poses a challenge of finding the reused text in a document. In this paper, three pre-retrieval query formulation strategies for heuristic retrieval in case of low obfuscated, high obfuscated, and translated text are studied. The strategies used are (a) Query formulation using proper nouns; (b) Query formulation using unique words (Hapax); and (c) Query formulation using most frequent words. Whereas in case of low and high obfuscation and simulated paraphrasing, keywords with Hapax proved to be slightly more efficient, initial results indicate that the simple strategy of query formulation using proper nouns gives promising results and may prove better in reducing the size of the corpus for post processing, for identifying local text reuse in case of obfuscated and translated text reuse.

• 국제학술발표논문집
• /
• The 6th International Conference on Construction Engineering and Project Management
• /
• pp.37-38
• /
• 2015

A number of documents containing important and useful knowledge have been generated over time in the construction industry. Such text-based knowledge plays an important role in the construction industry for decision-making and business strategy development by being used as best practice for upcoming projects, delivering lessons learned for better risk management and project control. Thus, practical and usable knowledge creation from construction documents is necessary to improve business efficiency. This study proposes a knowledge creating system from construction documents using text mining and the design comprises three main steps - text mining preprocessing, weight calculation of each term, and visualization. A system prototype was developed as a pilot study of the system design. This study is significant because it validates a knowledge creating system design based on text mining and visualization functionality through the developed system prototype. Automated visualization was found to significantly reduce unnecessary time consumption and energy for processing existing data and reading a range of documents to get to their core, and helped the system to provide an insight into the construction industry.

• 한국콘텐츠학회논문지
• /
• 제11권12호
• /
• pp.560-569
• /
• 2011

본 논문에서는 개방형 OS인 안드로이드에서 힌디어 텍스트 처리방법을 제안한다. 텍스트 처리의 핵심은 알파벳을 문자로 조합하는 규칙을 정의하는 오토마타와 폰트 파일에서 문자에 대응하는 이미지를 검색하고 이를 화면에 표시하는 폰트 렌더링이다. 오토마타는 입력 문자의 종류와 개수에 좌우되는데 유니코드를 기반으로 자음 14자와 모음 34자를 알파벳으로 사용하는 오토마타를 제안한다. 조합된 음절은 테이블 매핑 방식을 사용하여 그립 인덱스로 변환하고 해당하는 폰트를 로드하기 위한 핸들로 사용한다. 프리 타입 폰트엔진의 다국어 지원 프레임워크에 따라 제안방법을 별도의 모듈로 추가함으로서 시스템 수준에서 힌디어를 지원할 수 있다. 메시지 어플리케이션을 통해 제안방법의 타당성을 보인다.

• 한국콘텐츠학회논문지
• /
• 제15권9호
• /
• pp.425-435
• /
• 2015

유머를 유발하는 핵심요인이 불합치라는 것에서는 많은 연구자들의 견해가 일치하지만, 불합치를 유발하는 상이한 스키마들의 활성화에서는 다른 견해를 보인다. 스키마의 활성화와 관련된 견해들 중에서 대비되는 두 견해에는 동시 활성화 견해(concurrent activation view)와 선택적 활성화 견해(selective activation view)가 있다. 이 연구에서는 유머텍스트 처리에서 상이한 두 스키마들이 어떻게 활성화 되는지를 경험적으로 검증함으로써 대비되는 두 견해를 살펴보고자 하였다. 실험은 유머 텍스트를 기저-도입-불합치-해소의 4단계로 구분하여 시행하였고, 불합치는 존재하되 해소부분을 조작한 대조텍스트를 사용하였다. 실험에서는 문맥에 부합하는 스키마를 S1, 부합하지 않고 유머의 참 뜻을 나타내는 스키마를 S2라 명명하였다. 실험결과 불합치 단계에서 활성화된 스키마들이 해소 단계에서도 여전히 활성화되고, 불합치 단계에서의 S1, S2 모두 해소단계의 S1, S2와 통계적으로 유의한 차이를 보였다. 이는 불합치 단계에서는 하나의 스키마가 다른 스키마를 억제할 것이라고 가정하는 선택적 활성화 견해와는 모순되는 것이다. 이 연구의 결과는 유머 텍스트를 처리하는 과정에서 상이한 스키마들은 동시에 활성화됨을 시사한다.

• International Journal of Internet, Broadcasting and Communication
• /
• 제12권2호
• /
• pp.45-50
• /
• 2020

A Language model with neural networks commonly trained with likelihood loss. Such that the model can learn the sequence of human text. State-of-the-art results achieved in various language generation tasks, e.g., text summarization, dialogue response generation, and text generation, by utilizing the language model's next token output probabilities. Monotonous and boring outputs are a well-known problem of this model, yet only a few solutions proposed to address this problem. Several decoding techniques proposed to suppress repetitive tokens. Unlikelihood training approached this problem by penalizing candidate tokens probabilities if the tokens already seen in previous steps. While the method successfully showed a less repetitive generated token, the method has a large memory consumption because of the training need a big vocabulary size. We effectively reduced memory footprint by encoding words as sequences of subword units. Finally, we report competitive results with token level unlikelihood training in several automatic evaluations compared to the previous work.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 추계종합학술대회 논문집(3)
• /
• pp.121-124
• /
• 2000

In this paper, we propose a pen gesture recognition system for user interface in multimedia terminal which requires fast processing time and high recognition rate. It is realtime and interaction system between graphic and text module. Text editing in recognition system is performed by pen gesture in graphic module or direct editing in text module, and has all 14 editing functions. The pen gesture recognition is performed by searching classification features that extracted from input strokes at pen gesture model. The pen gesture model has been constructed by classification features, ie, cross number, direction change, direction code number, position relation, distance ratio information about defined 15 types. The proposed recognition system has obtained 98% correct recognition rate and 30msec average processing time in a recognition experiment.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2006년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.261-263
• /
• 2006

Nowadays, technologies such as RFID, sensor network makes our life comfortable more and more. In this paper we propose a wearable computing system for blind and deaf person who can be easily out of sight from our technology. We are making a wearable computing system that is consisted of embedded board to processing data, ultrasonic sensors to get distance data and motors that make vibration as a signal to see the screen for a deaf person. This system offers environmental informations by text and voice. For example, distance data from a obstacle to a person are calculated by data compounding module using sensed ultrasonic reflection time. This data is converted to text or voice by main processing module, and are serviced to a handicapped person. Furthermore we will extend this system using a voice recognition module and text to voice convertor module to help communication among the blind and deaf persons.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2016년도 추계학술발표대회
• /
• pp.529-532
• /
• 2016

본 논문은 기존의 TextRank 알고리즘에 상호정보량 척도를 결합하여 군집 기반에서 키워드 추출하는 ClusterTextRank 기법을 제안한다. 제안 기법은 k-means 군집화 알고리즘을 이용하여 문서들을 여러 군집으로 나누고, 각 군집에 포함된 단어들을 최소신장트리 그래프로 표현한 후 이에 근거한 군집 정보량을 고려하여 키워드를 추출한다. 제안 기법의 성능을 평가하기 위해 여행 관련 블로그 데이터를 이용하였으며, 제안 기법이 기존 TextRank 알고리즘보다 키워드 추출의 정확도가 약 13% 가량 개선됨을 보인다.

• 한국정보처리학회논문지
• /
• 제7권4호
• /
• pp.1150-1159
• /
• 2000

This paper presents an automatic text categorization model that improves the accuracy by combining statistical and knowledge-based categorization methods. In our model we apply knowledge-based method first, and then apply statistical method on the text which are not categorized by knowledge-based method. By using this combined method, we can improve the accuracy of categorization while categorize all the texts without failure. For statistical categorization, the vector model with Inverted Category Frequency (ICF) weighting is used. For knowledge-based categorization, Phrasal Patterns and Keyword Sets are introduced to represent sentence patterns, and then pattern matching is performed. Experimental results on new articles show that the accuracy of categorization can be improved by combining the tow different categorization methods.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2017년도 추계학술발표대회
• /
• pp.395-398
• /
• 2017

We propose a web crawler system with keyword extraction function in this paper. Researches on the keyword extraction in existing text mining are mostly based on databases which have already been grabbed by documents or corpora, but the purpose of this paper is to establish a real-time keyword extraction system which can extract the keywords of the corresponding text and store them into the database together while grasping the text of the web page. In this paper, we design and implement a crawler combining RAKE keyword extraction algorithm. It can extract keywords from the corresponding content while grasping the content of web page. As a result, the performance of the RAKE algorithm is improved by increasing the weight of the important features (such as the noun appearing in the title). The experimental results show that this method is superior to the existing method and it can extract keywords satisfactorily.