• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.5
• /
• pp.1396-1412
• /
• 2023

Conversation modeling is an important and challenging task in the field of natural language processing because it is a key component promoting the development of automated humanmachine conversation. Most recent research concerning conversation modeling focuses only on the current utterance (considered as the current question) to generate a response, and thus fails to capture the conversation's logic from its beginning. Some studies concatenate the current question with previous conversation sentences and use it as input for response generation. Another approach is to use an encoder to store all previous utterances. Each time a new question is encountered, the encoder is updated and used to generate the response. Our approach in this paper differs from previous studies in that we explicitly separate the encoding of the question from the encoding of its context. This results in different encoding models for the question and the context, capturing the specificity of each. In this way, we have access to the entire context when generating the response. To this end, we propose a deep neural network-based model, called the Context Model, to encode previous utterances' information and combine it with the current question. This approach satisfies the need for context information while keeping the different roles of the current question and its context separate while generating a response. We investigate two approaches for representing the context: Long short-term memory and Convolutional neural network. Experiments show that our Context Model outperforms a baseline model on both ConvAI2 Dataset and a collected dataset of conversational English.

• Journal of the Korean Society for Library and Information Science
• /
• v.51 no.4
• /
• pp.99-120
• /
• 2017

In order to develop a semiautomatic support system that allows researchers concerned to efficiently analyze the technical trends for the ever-growing industry and market. This paper introduces a couple of Korean sentence generation models that can automatically generate definitional statements as well as descriptions of technical terms and concepts. The proposed models are based on a deep learning model called LSTM (Long Sort-Term Memory) capable of effectively labeling textual sequences by taking into account the contextual relations of each item in the sequences. Our models take technical terms as inputs and can generate a broad range of heterogeneous textual descriptions that explain the concept of the terms. In the experiments using large-scale training collections, we confirmed that more accurate and reasonable sentences can be generated by CHAR-CNN-LSTM model that is a word-based LSTM exploiting character embeddings based on convolutional neural networks (CNN). The results of this study can be a force for developing an extension model that can generate a set of sentences covering the same subjects, and furthermore, we can implement an artificial intelligence model that automatically creates technical literature.

• 한국노년학
• /
• v.40 no.4
• /
• pp.659-689
• /
• 2020

This study focuses on the process of internalizing the perception of negative discrimination among elderly people who have experienced age discrimination. The grounded theory method was used to identify the age discrimination experienced by the elderly in their daily lives and to explore the consequences of such discrimination through self-ageism. According to the analysis results, the elderly respond not only to explicit discrimination but also to implicit discrimination. In this process, the stronger the pain of old age (poverty, disease, ignorance, and solitude) and the stronger the memory of failing to respond to discrimination, the higher the level of self-ageism. "Self-ageism" has internalized the negative image of the elderly, who have been discriminated against by age, resulting in disadvantages in terms of quality of life through various reject/separate/suspension actions in reality. In the presence of supportive resources such as health, institutional support, and caregivers, the elderly have room to overcome self-ageism through more active ways. However elderly people who cannot motivate themselves and they lack these resources, elderly are trapped in negative reflux caused by self-ageism and sustained a depressed and shrinking life. In this state, it is not easy to motivate themselves and make physical, cognitive, emotional and social responses. Therefore, in this analysis, we specifically noted two aspects, contextual conditions and interventional conditions, and proposed programs for synchronizing senior citizens and improving resiliency from a microscopic point of view, and argued for the need to develop systems such as supplementing welfare and health service systems related to the entire life cycle, expanding accessibility and 'age-integration' through 'Community Care', awareness improvement and anti-discrimination laws.

• The Korean Journal of Archival Studies
• /
• no.74
• /
• pp.45-78
• /
• 2022

South Africa implemented apartheid from 1948 to 1994. The main content of this policy was to classify races such as whites, Indians, mixed-race people, and blacks, and to limit all social activities, including residence, personal property ownership, and economic activities, depending on the class. All races except white people were discriminated against and suppressed for having different skin colors. South African citizens resisted the government's indiscriminate violence, and public opinion criticizing them expanded beyond the local community to various parts of the world. One of the things that made this possible was photographs detailing the scene of the violence. Foreign journalists who captured popular oppression as well as photographers from South Africa were immersed in recording the lives of those who were marginalized and suffered on an individual level. If they had not been willing to inform the reality and did not actually record it as a photo, many people would not have known the horrors of the situation caused by racial discrimination. Therefore, this paper focuses on Rise and Fall of Apartheid: Photography and the Bureau of Everyday Life, which captures various aspects of apartheid and displays related records, and examines the aspects of racism committed in South Africa described in the photo. The exhibition covers the period from 1948 when apartheid began until 1995, when Nelson Mandela was elected president and the Truth and Reconciliation Commission was launched to correct the wrong view of history. Many of the photos on display were taken by Peter Magubane, Ian Berry, David Goldblatt, and Santu Mofoken, a collection of museums, art galleries and media, including various archives. The photographs on display are primarily the work of photographers. It is both a photographic work and a media that proves South Africa's past since the 1960s, but it has been mainly dealt with in the field of photography and art history rather than from a historical or archival point of view. However, the photos have characteristics as records, and the contextual information contained in them is characterized by being able to look back on history from various perspectives. Therefore, it is very important to expand in the previously studied area to examine the time from various perspectives and interpret it anew. The photographs presented in the exhibition prove and describe events and people that are not included in South Africa's official records. This is significant in that it incorporates socially marginalized people and events into historical gaps through ordinary people's memories and personal records, and is reproduced in various media to strengthen and spread the context of record production.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.253-266
• /
• 2018

This paper proposes a methodology applying sequence tagging methodology to improve the performance of NER(Named Entity Recognition) used in QA system. In order to retrieve the correct answers stored in the database, it is necessary to switch the user's query into a language of the database such as SQL(Structured Query Language). Then, the computer can recognize the language of the user. This is the process of identifying the class or data name contained in the database. The method of retrieving the words contained in the query in the existing database and recognizing the object does not identify the homophone and the word phrases because it does not consider the context of the user's query. If there are multiple search results, all of them are returned as a result, so there can be many interpretations on the query and the time complexity for the calculation becomes large. To overcome these, this study aims to solve this problem by reflecting the contextual meaning of the query using Bidirectional LSTM-CRF. Also we tried to solve the disadvantages of the neural network model which can't identify the untrained words by using ontology knowledge based feature. Experiments were conducted on the ontology knowledge base of music domain and the performance was evaluated. In order to accurately evaluate the performance of the L-Bidirectional LSTM-CRF proposed in this study, we experimented with converting the words included in the learned query into untrained words in order to test whether the words were included in the database but correctly identified the untrained words. As a result, it was possible to recognize objects considering the context and can recognize the untrained words without re-training the L-Bidirectional LSTM-CRF mode, and it is confirmed that the performance of the object recognition as a whole is improved.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.37 no.2
• /
• pp.40-53
• /
• 2019

This study was conducted to investigate the changes in Scenic Spots by utilizing the "Sinjeungdonggukyeojiseungram(新增東國輿地勝覽)" and "Joseonhwanyeoseungnam(朝鮮?與勝覽)" to interpret "Joseonmyeongseungsiseon(朝鮮名勝詩選, 1915)". By examining the historical context when "Joseonmyeongseungsiseon" was published, it documented the Japanese's memories of 'Sino-Japanese War(淸日戰爭)' in 1894, which implies the 'policy of assimilation' by the Japanese Government-General of Korea after the Japanese annexation of Korea(1910). Detailed information about the author 'Narushima Sagimura(成島鷺村)' can be found in preface. In the "Joseonmyeongseungsiseon", it dedicates most of the part in describing the Scenic Spot especially in 'Anseong Do (15 lines)', where has the memories of war such as the 'First Sino-Japanese War'. The number of Scenic Spots, commonly mentioned in both "Sinjeungdonggukyeoji seungram" and "Joseonhwanyeoseungnam" in Cheonan province are 13 in total. Most of the content contains a similar structure. But 'Honggyeongwon(弘慶院)' and 'Seonghwanyeok(成歡驛)' has both the common Joseon Dynasty landscape point of view, and the additional historical context which are about the 'Jeongyujaeran(丁酉再亂)' or 'First Sino-Japanese War' consequently enlightens the 'placeness' of the Scenic Spots. Among the newly described Scenic Spots, 'Anseongdo(安城渡)' is the part that focuses on the memory of the 'Anseongcheon Battle' that gave Japan its first defeat in the Sino-Japanese War. Especially, by introducing the poetry of 'Sinobu Shunpei' it maximizes the appreciation through emphasizing the direct correlation between placeness and the poem itself. While the Joseon Dynasty poems are 10 pieces in total and their title and the subject matters are all related to historical spots, and the appreciation also maximizes when fully interpreted with understanding the historical context. However, it's contextual meanings are neglected by dividing the actual structures into separate pages. When looking at the location of famous historic sites, they come in many different types, considering the location, meaning, size, and conditions surrounding them. It appears as a service space for travelers, a place for sightseeing, relaxation or return, a temple space for paying respects or memorial services, a fortress facility for defense and protection, or a fishing area for wages, and an old battlefield. Especially, it is noted that the area is diverse as the cultural landscape of Cheonan, given that the battle space between the hermitage and the Donghangnongmin(東學農民) is shared with each other. It is necessary to establish policies for the preservation and restoration of local cultural assets based on these points in the future.