• The Journal of the Korea Contents Association
• /
• v.19 no.10
• /
• pp.620-629
• /
• 2019

An automated minutes summary system is required to objectively summarize and classify the contents of discussions or discussions for decision making. This paper designs and implements a minutes summary system using word2vec model to complement the existing minutes summary system. The proposed system is further implemented with word2vec model to remove index words during morpheme analysis and to extract representative sentences with common opinions from documents. The proposed system automatically classifies documents collected during the meeting process and extracts representative sentences representing the agenda among various opinions. The conference host can quickly identify and manage all the agendas discussed at the meeting through the proposal system. The proposed system analyzes various agendas of large-scale debates or discussions and summarizes sentences that can be representative opinions to support fast and accurate decision making.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.7
• /
• pp.949-955
• /
• 2022

Deep learning-based text summarization models are not free from datasets. For example, a summarization model trained with a news summarization dataset is not good at summarizing other types of texts such as internet posts and papers. In this study, we define this phenomenon as Data Bias Problem (DBP) and propose two training methods for solving it. The first is the 'proper nouns masking' that masks proper nouns. The second is the 'length variation' that randomly inflates or deflates the length of text. As a result, experiments show that our methods are efficient for solving DBP. In addition, we analyze the results of the experiments and present future development directions. Our contributions are as follows: (1) We discovered DBP and defined it for the first time. (2) We proposed two efficient training methods and conducted actual experiments. (3) Our methods can be applied to all summarization models and are easy to implement, so highly practical.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.8
• /
• pp.41-47
• /
• 2022

In this paper, we propose how to implement text summaries for colloquial data that are not clearly organized. For this study, SAMSum data, which is colloquial data, was used, and the BERTSumExtAbs model proposed in the previous study of the automatic summary model was applied. More than 70% of the SAMSum dataset consists of conversations between two people, and the remaining 30% consists of conversations between three or more people. As a result, by applying the automatic text summarization model to colloquial data, a result of 42.43 or higher was derived in the ROUGE Score R-1. In addition, a high score of 45.81 was derived by fine-tuning the BERTSum model, which was previously proposed as a text summarization model. Through this study, the performance of colloquial generation summary has been proven, and it is hoped that the computer will understand human natural language as it is and be used as basic data to solve various tasks.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.11
• /
• pp.501-512
• /
• 2021

Recently, a breakthrough has been made in the NLP area by Transformer techniques based on encoder-decoder. However, this only can be used in mainstream languages where millions of dataset are well-equipped, such as English and Chinese, and there is a limitation that it cannot be used in non-mainstream languages where dataset are not established. In addition, there is a deflection problem that focuses on the beginning of the document in mechanical summarization. Therefore, these methods are not suitable for documents with flows such as fairy tales and novels. In this paper, we propose a hybrid summarization method that does not require a dataset and improves the deflection problem using GAN with two adaptive discriminators. We evaluate our model on the CNN/Daily Mail dataset to verify an objective validity. Also, we proved that the model has valid performance in Korean, one of the non-mainstream languages.

• Journal of the Korean Society for Library and Information Science
• /
• v.56 no.1
• /
• pp.95-117
• /
• 2022

This study developed and evaluated audio-visual (AV) semantics-based video summarization methods using eye tracking and electroencephalography (EEG) data. For this study, twenty-seven university students participated in eye tracking and EEG experiments. The evaluation results showed that the average recall rate (0.73) of using both EEG and pupil diameter data for the construction of a video summary was higher than that (0.50) of using EEG data or that (0.68) of using pupil diameter data. In addition, this study reported that the reasons why the average recall (0.57) of the AV semantics-based personalized video summaries was lower than that (0.69) of the AV semantics-based generic video summaries. The differences and characteristics between the AV semantics-based video summarization methods and the text semantics-based video summarization methods were compared and analyzed.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.72-74
• /
• 2022

The National Science and Technology Knowledge Information Service (NTIS) provides information on national R&D projects. Project information consists of meta-information such as 'project name', 'project performance institution', 'research manager name', and text explaining projects such as 'research goal', 'research content', and 'expected effect'. There is a problem that it takes a lot of time to find the desired project information by checking all of the "research goals" or "research contents" in the list of results of searching for 1 million project information. To solve this problem, this paper proposes a project information summary system that summarizes the parts consisting of long texts within the national R&D project information. By analyzing the linguistic characteristics of the Korean language, a preprocessor was built and a project information summary model based on natural language processing technology was developed to process preprocessed text information. Through this, project information composed of long sentences is provided in a compressed and summarized form, which will help users to easily and quickly infer the overall content with the summary information alone.

• The KIPS Transactions:PartC
• /
• v.16C no.6
• /
• pp.725-730
• /
• 2009

Recently, there has been increasing interest to web access through mobile host due to the explosion of internet mobile terminal such as smart phone. However, small displays of mobile hosts make it difficult to browse the full content of a web page at a time. In order to overcome these limitation, we have designed and implemented Web-based text summarization system. The proposed system can summarize the text for the Web page in which abundant text exist in a page. This can reduce the amount of data transmission and minimize the unnecessary data output during browsing at mobile host. Through implementation, we have confirmed the functions of the proposed system.

• Journal of KIISE
• /
• v.44 no.5
• /
• pp.503-509
• /
• 2017

In this paper, the copy mechanism and input feeding are applied to recurrent neural network(RNN)-search model in a Korean-document summarization in an end-to-end manner. In addition, the performances of the document summarizations are compared according to the model and the tokenization format; accordingly, the syllable-unit, morpheme-unit, and hybrid-unit tokenization formats are compared. For the experiments, Internet newspaper articles were collected to construct a Korean-document summary data set (train set: 30291 documents; development set: 3786 documents; test set: 3705 documents). When the format was tokenized as the morpheme-unit, the models with the input feeding and the copy mechanism showed the highest performances of ROUGE-1 35.92, ROUGE-2 15.37, and ROUGE-L 29.45.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2006.05a
• /
• pp.473-476
• /
• 2006

인터넷의 발달로 인하여 정보의 양은 시간이 지날수록 폭발적으로 증가하고 있다. 이러한 방대한 정보로부터 정보검색시스템은 사용자에게 너무 많은 검색결과를 제시하여 사용자가 원하는 정보를 찾기 위해 너무 많은 시간을 소요하게 하는 정보의 과적재 문제가 있다. 질의 기반의 문서요약은 정보의 사용자가 원하는 정보의 검색시간을 줄임으로써 정보의 과적재 문제를 해결하는 방법으로서 점차 중요성이 증가하고 있다. 본 논문은 비음수 행렬 인수분해 (NMF, Non-negative Matrix Factorization)과 코사인 유사도를 이용하여 질의 기반의 문서를 요약하는 새로운 방법을 제안하였다. 제안된 방법은 질의와 문서 간에 사전학습이 필요 없다. 또한 문서를 그래프로 변형시키는 복잡한 처리 없이 NMF 에 의해 얻어진 의미 특징(semantic feature)과 의미 변수(semantic variable)로 문서의 고유 구조를 반영하여 요약의 정확도를 높일 수 있다. 마지막으로 단순한 방법으로 문장을 쉽게 요약할 수 있다.

• Journal of Internet Computing and Services
• /
• v.9 no.3
• /
• pp.169-178
• /
• 2008

In this paper, we propose a video summarization system which is based on activity in video acquired by multiple non-overlapping cameras for wide-area surveillance. The proposed system separates persons by time-independent background removal and detects activities of the segmented persons by their motions. In this paper, we extract eleven activities based on whose direction the persons move to and consider a key-frame as a frame which contains a meaningful activity. The proposed system summarizes based on activity-based key-frames and controls an amount of summarization according to an amount of activities. Thus the system can summarize videos by camera, time, and activity.