• Journal of Information Processing Systems
• /
• v.14 no.5
• /
• pp.1254-1271
• /
• 2018

Keyphrase extraction is one of fundamental natural language processing (NLP) tools to improve many text-mining applications such as document summarization and clustering. In this paper, we propose to use two novel techniques on the top of the state-of-the-art keyphrase extraction methods. First is the anti-patterns that aim to recognize non-keyphrase candidates. The state-of-the-art methods often used the rich feature set to identify keyphrases while those rich feature set cover only some of all keyphrases because keyphrases share very few similar patterns and stylistic features while non-keyphrase candidates often share many similar patterns and stylistic features. Second one is to use the dependency graph instead of the word co-occurrence graph that could not connect two words that are syntactically related and placed far from each other in a sentence while the dependency graph can do so. In experiments, we have compared the performances with different settings of the graphs (co-occurrence and dependency), and with the existing method results. Finally, we discovered that the combination method of dependency graph and anti-patterns outperform the state-of-the-art performances.

• Journal of Intelligence and Information Systems
• /
• v.24 no.4
• /
• pp.111-136
• /
• 2018

In this paper, we propose a methodology to extract answer information about queries from various types of unstructured documents collected from multi-sources existing on web in order to expand knowledge base. The proposed methodology is divided into the following steps. 1) Collect relevant documents from Wikipedia, Naver encyclopedia, and Naver news sources for "subject-predicate" separated queries and classify the proper documents. 2) Determine whether the sentence is suitable for extracting information and derive the confidence. 3) Based on the predicate feature, extract the information in the proper sentence and derive the overall confidence of the information extraction result. In order to evaluate the performance of the information extraction system, we selected 400 queries from the artificial intelligence speaker of SK-Telecom. Compared with the baseline model, it is confirmed that it shows higher performance index than the existing model. The contribution of this study is that we develop a sequence tagging model based on bi-directional LSTM-CRF using the predicate feature of the query, with this we developed a robust model that can maintain high recall performance even in various types of unstructured documents collected from multiple sources. The problem of information extraction for knowledge base extension should take into account heterogeneous characteristics of source-specific document types. The proposed methodology proved to extract information effectively from various types of unstructured documents compared to the baseline model. There is a limitation in previous research that the performance is poor when extracting information about the document type that is different from the training data. In addition, this study can prevent unnecessary information extraction attempts from the documents that do not include the answer information through the process for predicting the suitability of information extraction of documents and sentences before the information extraction step. It is meaningful that we provided a method that precision performance can be maintained even in actual web environment. The information extraction problem for the knowledge base expansion has the characteristic that it can not guarantee whether the document includes the correct answer because it is aimed at the unstructured document existing in the real web. When the question answering is performed on a real web, previous machine reading comprehension studies has a limitation that it shows a low level of precision because it frequently attempts to extract an answer even in a document in which there is no correct answer. The policy that predicts the suitability of document and sentence information extraction is meaningful in that it contributes to maintaining the performance of information extraction even in real web environment. The limitations of this study and future research directions are as follows. First, it is a problem related to data preprocessing. In this study, the unit of knowledge extraction is classified through the morphological analysis based on the open source Konlpy python package, and the information extraction result can be improperly performed because morphological analysis is not performed properly. To enhance the performance of information extraction results, it is necessary to develop an advanced morpheme analyzer. Second, it is a problem of entity ambiguity. The information extraction system of this study can not distinguish the same name that has different intention. If several people with the same name appear in the news, the system may not extract information about the intended query. In future research, it is necessary to take measures to identify the person with the same name. Third, it is a problem of evaluation query data. In this study, we selected 400 of user queries collected from SK Telecom 's interactive artificial intelligent speaker to evaluate the performance of the information extraction system. n this study, we developed evaluation data set using 800 documents (400 questions * 7 articles per question (1 Wikipedia, 3 Naver encyclopedia, 3 Naver news) by judging whether a correct answer is included or not. To ensure the external validity of the study, it is desirable to use more queries to determine the performance of the system. This is a costly activity that must be done manually. Future research needs to evaluate the system for more queries. It is also necessary to develop a Korean benchmark data set of information extraction system for queries from multi-source web documents to build an environment that can evaluate the results more objectively.

• Journal of Internet Computing and Services
• /
• v.18 no.6
• /
• pp.55-63
• /
• 2017

One of the purposes of Information Technology (IT) is to support human response to natural and social problems such as natural disasters and spread of disease, and to improve the quality of human life. Recent climate change has happened worldwide, natural disasters threaten the quality of life, and human safety is no longer guaranteed. IT must be able to support tasks related to disaster response, and more importantly, it should be used to predict and minimize future damage. In South Korea, the data related to the damage is checked out by each local government and then federal government aggregates it. This data is included in disaster reports that the federal government discloses by disaster case, but it is difficult to obtain raw data of the damage even for research purposes. In order to obtain data, information extraction may be applied to disaster reports. In the field of information extraction, most of the extraction targets are web documents, commercial reports, SNS text, and so on. There is little research on information extraction for government disaster reports. They are mostly text, but the structure of each sentence is very different from that of news articles and commercial reports. The features of the government disaster report should be carefully considered. In this paper, information extraction method for South Korea government reports in the word format is presented. This method is based on patterns and dictionaries and provides some additional ideas for tokenizing the damage representation of the text. The experiment result is F1 score of 80.2 on the test set. This is close to cutting-edge information extraction performance before applying the recent deep learning algorithms.

• Journal of Intelligence and Information Systems
• /
• v.25 no.3
• /
• pp.161-177
• /
• 2019

In this paper, we study the performance improvement of the answer extraction in Question-Answering system by using sentence dependency parsing result. The Question-Answering (QA) system consists of query analysis, which is a method of analyzing the user's query, and answer extraction, which is a method to extract appropriate answers in the document. And various studies have been conducted on two methods. In order to improve the performance of answer extraction, it is necessary to accurately reflect the grammatical information of sentences. In Korean, because word order structure is free and omission of sentence components is frequent, dependency parsing is a good way to analyze Korean syntax. Therefore, in this study, we improved the performance of the answer extraction by adding the features generated by dependency parsing analysis to the inputs of the answer extraction model (Bidirectional LSTM-CRF). The process of generating the dependency graph embedding consists of the steps of generating the dependency graph from the dependency parsing result and learning the embedding of the graph. In this study, we compared the performance of the answer extraction model when inputting basic word features generated without the dependency parsing and the performance of the model when inputting the addition of the Eojeol tag feature and dependency graph embedding feature. Since dependency parsing is performed on a basic unit of an Eojeol, which is a component of sentences separated by a space, the tag information of the Eojeol can be obtained as a result of the dependency parsing. The Eojeol tag feature means the tag information of the Eojeol. The process of generating the dependency graph embedding consists of the steps of generating the dependency graph from the dependency parsing result and learning the embedding of the graph. From the dependency parsing result, a graph is generated from the Eojeol to the node, the dependency between the Eojeol to the edge, and the Eojeol tag to the node label. In this process, an undirected graph is generated or a directed graph is generated according to whether or not the dependency relation direction is considered. To obtain the embedding of the graph, we used Graph2Vec, which is a method of finding the embedding of the graph by the subgraphs constituting a graph. We can specify the maximum path length between nodes in the process of finding subgraphs of a graph. If the maximum path length between nodes is 1, graph embedding is generated only by direct dependency between Eojeol, and graph embedding is generated including indirect dependencies as the maximum path length between nodes becomes larger. In the experiment, the maximum path length between nodes is adjusted differently from 1 to 3 depending on whether direction of dependency is considered or not, and the performance of answer extraction is measured. Experimental results show that both Eojeol tag feature and dependency graph embedding feature improve the performance of answer extraction. In particular, considering the direction of the dependency relation and extracting the dependency graph generated with the maximum path length of 1 in the subgraph extraction process in Graph2Vec as the input of the model, the highest answer extraction performance was shown. As a result of these experiments, we concluded that it is better to take into account the direction of dependence and to consider only the direct connection rather than the indirect dependence between the words. The significance of this study is as follows. First, we improved the performance of answer extraction by adding features using dependency parsing results, taking into account the characteristics of Korean, which is free of word order structure and omission of sentence components. Second, we generated feature of dependency parsing result by learning - based graph embedding method without defining the pattern of dependency between Eojeol. Future research directions are as follows. In this study, the features generated as a result of the dependency parsing are applied only to the answer extraction model in order to grasp the meaning. However, in the future, if the performance is confirmed by applying the features to various natural language processing models such as sentiment analysis or name entity recognition, the validity of the features can be verified more accurately.

• Journal of the Korea Convergence Society
• /
• v.9 no.11
• /
• pp.23-28
• /
• 2018

Recent approaches to Relation Extraction methods mostly tend to be limited to mention level relation extractions. These types of methods, while featuring high performances, can only extract relations limited to a single sentence or so. The inability to extract these kinds of data is a terrible amount of information loss. To tackle this problem this paper presents an Augmented External Memory Neural Network model to enable Global Relation Extraction. the proposed model's Global relation extraction is done by first gathering and analyzing the mention level relation extraction by the Augmented External Memory. Additionally the proposed model shows high level of performances in korean due to the fact it can take the often omitted subjects and objectives into consideration.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.5
• /
• pp.662-667
• /
• 2022

Semantic role labeling task(SRL) is to extract predicate and arguments such as agent, patient, place, time. In the previously SRL task studies, a pipeline method extracting linguistic features of sentence has been proposed, but in this method, errors of each extraction work in the pipeline affect semantic role labeling performance. Therefore, methods using End-to-End neural network model have recently been proposed. In this paper, we propose a neural network model using the Biaffine Average Attention model for SRL task. The proposed model consists of a structure that can focus on the entire sentence information regardless of the distance between the predicate in the sentence and the arguments, instead of LSTM model that uses the surrounding information for prediction of a specific token proposed in the previous studies. For evaluation, we used F1 scores to compare two models based BERT model that proposed in existing studies using F1 scores, and found that 76.21% performance was higher than comparison models.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.05a
• /
• pp.397-400
• /
• 2021

To solve the problem that existing computing methods cannot adequately represent the semantic features of sentences, Siamese TRAT, a semantic feature extraction model based on Transformer encoder is proposed. The transformer model is used to fully extract the semantic information within sentences and carry out deep semantic coding for sentences. In addition, the interactive attention mechanism is introduced to extract the similar features of the association between two sentences, which makes the model better at capturing the important semantic information inside the sentence. As a result, it improves the semantic understanding and generalization ability of the model. The experimental results show that the proposed model can improve the accuracy significantly for the semantic similarity calculation task of English and Chinese, and is more effective than the existing methods.

• The KIPS Transactions:PartB
• /
• v.14B no.1 s.111
• /
• pp.33-42
• /
• 2007

This paper presents a new methodology for music structure analysis which facilitates rhetoric-based music summarization. Similarity analysis of musical constituents suggests the structure of a musical piece. We can recognize its musical form from the structure. Musical forms have rhetorical characteristics of their on. We have utilized the characteristics for locating musical motifs. Motif extraction is to music summarization what topic sentence extraction is to text summarization. We have evaluated the effectiveness of this methodology through a popular music case study.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2000.12a
• /
• pp.181-184
• /
• 2000

The new method of feature extraction is proposed, considering the statistic feature of human voice, unlike the conventional methods of voice extraction. PCA(principal Component Analysis) is applied to this new method. PCA removes the repeating of data after finding the axis direction which has the greatest variance in input dimension. Then the new method is applied to real voice recognition to assess performance. When results of the number recognition in this paper and the conventional Mel-Cepstrum of voice feature parameter are compared, there is 0.5% difference of recognition rate. Better recognition rate is expected than word or sentence recognition in that less convergence time than the conventional method in extracting voice feature. Also, better recognition tate is expected when the optimum vector is used by statistic feature of data.

• Journal of Information Processing Systems
• /
• v.19 no.6
• /
• pp.842-857
• /
• 2023

Currently, most sentiment classification models on microblogging platforms analyze sentence parts of speech and emoticons without comprehending users' emotional inclinations and grasping moral nuances. This study proposes a hybrid sentiment analysis model. Given the distinct nature of microblog comments, the model employs a combined stop-word list and word2vec for word vectorization. To mitigate local information loss, the TextCNN model, devoid of pooling layers, is employed for local feature extraction, while BiLSTM is utilized for contextual feature extraction in deep learning. Subsequently, microblog comment sentiments are categorized using a classification layer. Given the binary classification task at the output layer and the numerous hidden layers within BiLSTM, the Tanh activation function is adopted in this model. Experimental findings demonstrate that the enhanced TextCNN-BiLSTM model attains a precision of 94.75%. This represents a 1.21%, 1.25%, and 1.25% enhancement in precision, recall, and F1 values, respectively, in comparison to the individual deep learning models TextCNN. Furthermore, it outperforms BiLSTM by 0.78%, 0.9%, and 0.9% in precision, recall, and F1 values.