• Korean Journal of Cognitive Science
• /
• v.2 no.2
• /
• pp.405-428
• /
• 1990

Language has a social function to communicate information. Linguists have gradually paid their attention to the function of language since the nineteen sixties, especially to the relationship of form, meaning and the function. The relationship could be more clearly grasped through disciyrse-based analysis than through sentence-based analysis. Many researches were centered on the discourse functional notion of topic. In the early 1970's the subject was defined as the grammatiocalized topic the topic as a discrete single constituent of the clause. In the late 1970's several lingusts including Givon suggerted that the topic was not an atomic, disctete entity, and that the clause could have more than one topic. The purpose of the present study is, following Givon, to study grammatical coding devices of topic and to measure the relative topic continuity/discontinuity of participant argu, ents in Korean narratives. By so doing, I would like to shed some light on effective ways of communicating information. The grammatical coding devices analyzed are the following eight structures: zero-anaphora, personal pronous, demonstrative pronouns, names, noun phrases following demonstratives, noun phrases following possessives, definite noun phrases and indefinite referentials. The narrative studied for the count was taken from the KoreanCIA chief's Testiomny:Revolution and Idol by Hyung Wook Kim. It was chosen because it was assumed that Kim's purpose in the novel was to tell a true story, which would not distort the natural use of language for literary effect. The measures taken in the analysis wre those of 'lookback', 'persistence', ambiguity'. The first of these, 'lookback', is a measure of the size of gap between the previous occurrence of a referent and its current occurence in the clause. The meausure of persistence, which is a measure of the speaker's topocal intent, reflects the topic's importance in the discourse. The third measure is a measure of ambiguity. This is necessary for assessing the disruptive effects that other topics within five previous clauses may have on topic identification. The more other topics are present within five previous clauses, the more difficult is the task of correct identification of a topic. The results of the present study show that the humanness of entities is the most powerful factior in topic continutiy in narrative discourse. The semantic roles of human arguments in narrative discourse tend to be agents or experiences. Since agents and experiences have high topicality in discourse, human entities clearly become clausal or discoursal topics. The results also show that the grammatical devices signal varying degrees of topic continuity discontinuity in continuous discourse. The more continuous a topic argument is, the less it is coded. For example, personal pronouns have the most continutiy and indefinite referentials have the least continutiy. The study strongly shows that topic continuity discontinutiy is controlled not only by grammatical devices available in the language but by socio-cultural factors and writer's intentions.

• Journal of KIISE:Software and Applications
• /
• v.41 no.8
• /
• pp.580-585
• /
• 2014

This paper conducts a variety of experiments for "the extraction of Korean parallel sentences using Wikipedia data". We refer to various methods that were previously proposed for other languages. We use two approaches. The first one is to use translation probabilities that are extracted from the existing resources such as Sejong parallel corpus, and the second one is to use dictionaries such as Wiki dictionary consisting of Wikipedia titles and MRDs (machine readable dictionaries). Experimental results show that we obtained a significant improvement in system using Wikipedia data in comparison to one using only the existing resources. We finally achieve an outstanding performance, an F1-score of 57.6%. We additionally conduct experiments using a topic model. Although this experiment shows a relatively lower performance, an F1-score of 51.6%, it is expected to be worthy of further studies.

• Proceedings of the Korean Society for Language and Information Conference
• /
• 2002.02a
• /
• pp.391-403
• /
• 2002

I argue that the so-called psychological predicates like komapta ′thankful,′ mwusepta ′fearful,′ silhta ′loathsome,′ or kulipta ′missing′require a nominative subject and a locative or dative complement, challenging the claim, a conventional wisdom originated from Kuno(1973), that they are two-place "transitive adjectives" requiring a nominative direct object, I also show that those adjectives are subject to having the locative-dative complement extracted, which is ultimately realized as a focused subject or a topic. Thus, in this type of double nominative constructions, the first nominative is a focused subject, and the second nominative forms an embedded clause with the psychological predicate, which functions as the predicate of the whole sentence.

• Speech Sciences
• /
• v.11 no.2
• /
• pp.89-104
• /
• 2004

This paper investigates previous theories on English focus realization and attempts to find out the overall acoustic characteristics of English focus. It has been argued in previous studies that English focus can be defined as a new information that is not recoverable from the context (Halliday 1967), a complementary element of presupposition (Jackendoff 1972), and what is predicated about the topic in a sentence (Sgall 1973, Gundel 1974). The phonetic realization of English focus in an utterance has been said to be either L+H*/H*, or falling accent. Yet it is a more or less simplified pattern not based on real data obtained from native speakers of English, and it does not consider the various pragmatic and contextual situations. In our experiments we found that native speakers uttered English focus sentences in different ways according to the different focus structure. Another notable result is that Korean speakers, when provided with the same experimental material, are neither able to distinguish different focus types nor deaccent the elements that are not focused in an utterance.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.1
• /
• pp.17-24
• /
• 2018

With improvement of speech recognition and natural language processing, dialog systems are recently adapted to various service domains. It became possible to get desirable services by conversation through the dialog system, but it is still necessary to improve separate modules, such as domain detection, intention detection, named entity recognition, and out-of-domain detection, in order to achieve stable service offer. When it misclassifies an in-domain sentence of conversation as out-of-domain, it will result in poor customer satisfaction and finally lost business. As there have been relatively small number of studies related to the out-of-domain detection, in this paper, we introduce a new method using a hierarchical Dirichlet process and demonstrate the effectiveness of it by experimental results on Korean dataset.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.15 no.1
• /
• pp.87-97
• /
• 2019

Reaction of people is importantly considered about specific case as a social network service grows. In the previous research on analysis of social network service, they predicted tendency of interesting topic by giving scores to sentences written by user. Based on previous study we proceeded research of sentiment analysis for social network service's sentences, which predict the result as positive or negative for movie reviews. In this study, we used movie review to get high accuracy. We classify the movie review into positive or negative based on the score for learning. Also, we performed embedding and morpheme analysis on movie review. We could predict learning result as positive or negative with a number 0 and 1 by applying the model based on learning result to social network service. Experimental result show accuracy of about 80% in predicting sentence as positive or negative.

• Journal of the Korea Convergence Society
• /
• v.12 no.7
• /
• pp.193-205
• /
• 2021

The AI speaker market is growing steadily. However, the satisfaction of actual users is only 42%. Therefore, in this paper, we collected reviews on Amazon Echo Dot 3rd and 4th generation models to analyze what hinders the user experience through the topic changes and emotional changes of each generation of AI speakers. By using topic modeling analysis techniques, we found changes in topics and topics that make up reviews for each generation, and examined how user sentiment on topics changed according to generation through deep learning-based sentiment analysis. As a result of topic modeling, five topics were derived for each generation. In the case of the 3rd generation, the topic representing general features of the speaker acted as a positive factor for the product, while user convenience features acted as negative factor. Conversely, in the 4^th generation, general features were negatively, and convenience features were positively derived. This analysis is significant in that it can present analysis results that take into account not only lexical features but also contextual features of the entire sentence in terms of methodology.

• Journal of Digital Contents Society
• /
• v.15 no.4
• /
• pp.559-568
• /
• 2014

Previous polarity classification using sentiment analysis utilizes a sentence rule by product reviews based rating points. It is difficult to be applied to blogs which have not rating of product reviews and is possible to fabricate product reviews by comment part-timers and managers who use web site so it is not easy to understand a product and store reviews which are reliability. Considering to these problems, if we analyze blogs which have personal and frank opinions and classify polarity, it is possible to understand rightly opinions for the product, store. This paper suggests that we extract high frequency vocabularies in blogs by several domains and choose topic words. Then we apply a technique of sentiment analysis and classify polarity about contents of blogs. To evaluate performances of sentiment analysis, we utilize the measurement index that use Precision, Recall, F-Score in an information retrieval field. In a result of evaluation, using suggested sentiment analysis is the better performances to classify polarity than previous techniques of using the sentence rule based product reviews.

• Journal of KIISE:Software and Applications
• /
• v.33 no.2
• /
• pp.243-251
• /
• 2006

In multi-document summarization by means of salient sentence extraction, it is important to remove redundant information. In the removal process, the similarities and differences of sentences are considered. In this paper, we propose a method for multi-document summarization which extracts salient sentences without having redundant sentences by way of cohesive term clustering method that utilizes co-occurrence Information. In the cohesive term clustering method, we assume that each term does not exist independently, but rather it is related to each other in meanings. To find the relations between terms, we cluster sentences according to topics and use the co-occurrence information oi terms in the same topic. We conduct experimental tests with the DUC(Document Understanding Conferences) data. In the tests, our method shows better performance of summarization than other summarization methods which use term co-occurrence information based on term cohesion of document or sentence unit, and simple statistical information.

• International Journal of Computer Science & Network Security
• /
• v.21 no.9
• /
• pp.281-291
• /
• 2021

The Islamic jurisprudential legal system represents an essential component of the Islamic religion, that governs many aspects of Muslims' daily lives. This creates many questions that require interpretations by qualified specialists, or Muftis according to the main sources of legislation in Islam. The Islamic jurisprudence is usually classified into branches, according to which the questions can be categorized and classified. Such categorization has many applications in automated question-answering systems, and in manual systems in routing the questions to a specialized Mufti to answer specific topics. In this work we tackle the problem of automatic categorisation of Islamic jurisprudential legal questions using deep learning techniques. In this paper, we build a hierarchical deep learning model that first extracts the question text features at two levels: word and sentence representation, followed by a text classifier that acts upon the question representation. To evaluate our model, we build and release the largest publicly available dataset of Islamic questions and answers, along with their topics, for 52 topic categories. We evaluate different state-of-the art deep learning models, both for word and sentence embeddings, comparing recurrent and transformer-based techniques, and performing extensive ablation studies to show the effect of each model choice. Our hierarchical model is based on pre-trained models, taking advantage of the recent advancement of transfer learning techniques, focused on Arabic language.