Award type：Award from Japanese society, conference, symposium, etc. 

Award type：Award from Japanese society, conference, symposium, etc. 

Award type：Award from international society, conference, symposium, etc.  Country：Netherlands

Award type：Award from Japanese society, conference, symposium, etc. 

Award type：Award from international society, conference, symposium, etc.  Country：Belgium

Award type：Award from Japanese society, conference, symposium, etc.  Country：Japan

Award type：Award from Japanese society, conference, symposium, etc.  Country：Japan

Award type：Award from Japanese society, conference, symposium, etc.  Country：Japan

CiNii Research

Publishing type：Research paper (international conference proceedings)   Publisher：Springer  

DOI： 10.1007/978-3-030-86475-0_18

Web of Science

Scopus

Other Link： https://dblp.uni-trier.de/db/conf/dexa/dexa2021-2.html#Komamizu21

Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/MIPR51284.2021.00018

Scopus

Other Link： https://dblp.uni-trier.de/db/conf/mipr/mipr2021.html#KomamizuIOT21

Publisher：WAT 2021 - 8th Workshop on Asian Translation, Proceedings of the Workshop  

For updating the translations of Japanese statutes based on their amendments, we need to consider the translation “focality;” that is, we should only modify expressions that are relevant to the amendment and retain the others to avoid misconstruing its contents. In this paper, we introduce an evaluation metric and a corpus to improve focality evaluations. Our metric is called an Inclusive Score for DIfferential Translation: (ISDIT). ISDIT consists of two factors: (1) the n-gram recall of expressions unaffected by the amendment and (2) the n-gram precision of the output compared to the reference. This metric supersedes an existing one for focality by simultaneously calculating the translation quality of the changed expressions in addition to that of the unchanged expressions. We also newly compile a corpus for Japanese partially amendment translation that secures the focality of the post-amendment translations, while an existing evaluation corpus does not. With the metric and the corpus, we examine the performance of existing translation methods for Japanese partially amendment translations.

Scopus

Publishing type：Research paper (scientific journal)  

DOI： 10.26421/JDI2.4-5

Publishing type：Research paper (scientific journal)  

DOI： 10.1007/s10115-020-01445-4

Web of Science

Scopus

Language：Japanese   Publishing type：Research paper (scientific journal)   Publisher：The Japanese Society for Artificial Intelligence  

<p>We propose a method that assists legislation drafters in finding inappropriate use of Japanese legal terms and their corrections from Japanese statutory sentences. In particular, we focus on sets of similar legal terms whose usages are strictly defined in legislation drafting rules that have been established over the years. In this paper, we first define input and output of legal term correction task. We regard it as a special case of sentence completion test with multiple choices. Next, we describe a legal term correction method for Japanese statutory sentences. Our method predicts suitable legal terms using Random Forest classifiers. The classifiers in our method use adjacent words to a target legal term as input features, and are optimized in various parameters including the number of adjacent words to be used for each legal term set. We conduct an experiment using actual statutory sentences from 3,983 existing acts and cabinet orders that consist of approximately 47M words in total. As for legal term sets, we pick 27 sets from legislation drafting manuals. The experimental result shows that our method outperformed existing modern word prediction methods using neural language models and that each Random Forest classifier utilizes characteristics of its corresponding legal term set.</p>

DOI： 10.1527/tjsai.h-j53

Scopus

CiNii Research

Language：English   Publisher：The Japanese Society for Artificial Intelligence  

<p>We propose a method to assist legislative drafters that locates inappropriate legal terms in Japanese statutorysentences and suggests corrections. We focus on sets of mistakable legal terms whose usages are defined in legislationdrafting rules. Our method predicts suitable legal terms using a classifier based on BERT (Bidirectional EncoderRepresentations from Transformers). The BERT classifier is pretrained with a huge number of whole sentences; thus,it contains abundant linguistic knowledge. Classifiers for predicting legal terms suffer from two-level infrequency:term-level infrequency and set-level infrequency. The former causes a class imbalance problem and the latter causesan underfitting problem; both degrade classification performance. To overcome these problems, we apply threetechniques, namely, preliminary domain adaptation, repetitive soft undersampling, and classifier unification. Thepreliminary domain adaptation improves overall performance by providing prior knowledge of statutory sentences,the repetitive soft undersampling overcomes term-level infrequency, and the classifier unification overcomes set-levelinfrequency while saving storage consumption. Our experiments show that our classifier outperforms conventionalclassifiers using Random Forest or language models, and that all three training techniques improve performance.</p>

DOI： 10.1527/tjsai.E-K25

Scopus

CiNii Research

Publisher：The Japanese Society for Artificial Intelligence  

<p>Legal documents contain legal terms that have similar meaning or pronunciation each other. Japanese legislation defines their usage on the basis of traditional customs and rules. In accordance with the definition, we need to use these legal terms properly and strictly in a statute. We are also encouraged to follow the definition in writing broad-sense legal documents, such as contracts and terms of use. To assist in writing legal documents, we propose a method that locates inappropriate legal terms in Japanese statutory sentences and suggests corrections. We solve this task with a classifier by regarding the task as a sentence completion test. Our classifier is based on a pretrained BERT model trained by using a large amount of general sentences. To raise performance, we apply three training techniques: domain adaptation, undersampling, classifier unification. Our experiments show that our classifier achieved better performance than Random Forest-based ones and language model-based ones.</p>

DOI： 10.11517/pjsai.JSAI2020.0_4P3OS805

Language：English   Publishing type：Research paper (other academic)   Publisher：Springer  

DOI： 10.1007/978-3-030-58790-1_1

Web of Science

Scopus

Other Link： https://dblp.uni-trier.de/db/conf/jsai/jsai2019w.html#KomamizuFOT19

Publishing type：Research paper (scientific journal)  

Other Link： https://dblp.uni-trier.de/db/journals/corr/corr2009.html#abs-2009-06194

Publisher：CEUR Workshop Proceedings  

Linked Open Data (LOD) has been proliferated over various domains, however, there are still lots of open data in various format other than RDF. Document-centric XML data are such open data that are connected with entities in LOD as supplemental documents for these entities. To utilize document-centric XML data linked from entities in LOD, in this paper, a SPARQL-based seamless access method on RDF and XML data is proposed. In particular, an extension to SPARQL, XQueryFILTER, which enables XQuery as a filter in SPARQL is proposed. For efficient query processing of the combination of SPARQL and XQuery, a query optimization is proposed. Experimental scenarios using real-world data showcase the effectiveness of XQueryFILTER and optimization efficiency.

Scopus

Publishing type：Research paper (international conference proceedings)   Publisher：Springer  

DOI： 10.1007/978-3-030-59051-2_14

Web of Science

Scopus

Other Link： https://dblp.uni-trier.de/db/conf/dexa/dexa2020-2.html#KomamizuUOT20

Language：English  

Language：English   Publishing type：Research paper (other academic)   Publisher：IEEE  

DOI： 10.1109/BigData47090.2019.9006511

Scopus

Other Link： https://dblp.uni-trier.de/db/conf/bigdataconf/bigdataconf2019.html#YamakoshiKOT19

Authorship：Lead author   Language：English  

Language：Japanese   Publishing type：Research paper (other academic)  

Authorship：Lead author   Language：English  

Language：Japanese   Publishing type：Research paper (other academic)  

Language：English  

Language：Japanese  

Language：Japanese   Publishing type：Research paper (other academic)  

Language：English  

Language：Japanese   Publishing type：Research paper (scientific journal)   Publisher：The Japanese Society for Artificial Intelligence  

Our purpose is to provide an automatic summarization for Japanese acts and we propose a sententence extraction method with Random Forest. While the traditional automatic summarization methods have used the information of summarizing source data, in recent years, the methods based on machine learning use the summarization results. However, in such a method, the amount of learning corpus is small, especially in Japanese text. In this research, we solve this problem by using "Outlines of Japanese Statutes," which are official summaries of statutes published by the Japanese government. Furthermore, we show that the sentence extraction method with Random Forest has higher performance rather than with decision trees or with support vector machines.

DOI： 10.11517/pjsai.jsai2019.0_4e2os7a02

CiNii Research

Language：English   Publishing type：Research paper (scientific journal)   Publisher：The Institute of Electronics, Information and Communication Engineers  

<p>Multi-attributed graphs, in which each node is characterized by multiple types of attributes, are ubiquitous in the real world. Detection and characterization of communities of nodes could have a significant impact on various applications. Although previous studies have attempted to tackle this task, it is still challenging due to difficulties in the integration of graph structures with multiple attributes and the presence of noises in the graphs. Therefore, in this study, we have focused on clusters of attribute values and strong correlations between communities and attribute-value clusters. The graph clustering methodology adopted in the proposed study involves <u><b>C</b></u>ommunity detection, <u><b>A</b></u>ttribute-value clustering, and deriving <u><b>R</b></u>elationships between communities and attribute-value clusters (CAR for short). Based on these concepts, the proposed multi-attributed graph clustering is modeled as CAR-clustering. To achieve CAR-clustering, a novel algorithm named CARNMF is developed based on non-negative matrix factorization (NMF) that can detect CAR in a cooperative manner. Results obtained from experiments using real-world datasets show that the CARNMF can detect communities and attribute-value clusters more accurately than existing comparable methods. Furthermore, clustering results obtained using the CARNMF indicate that CARNMF can successfully detect informative communities with meaningful semantic descriptions through correlations between communities and attribute-value clusters.</p>

DOI： 10.1587/transinf.2018DAP0022

Web of Science

Scopus

CiNii Research

Web of Science

Publisher：CEUR Workshop Proceedings  

As law is a blueprint of a society and is changed over time as social environments changed, analyzing histories (change provenances) of laws can reveal important facts such as legislative facts and critical events for the society. Linked Open Data (LOD) has emerged as a preferred method for publishing and sharing open data, however, there is an ontological barrier for publishing law history data as LOD. To break through the barrier, this paper proposes an ontology for law history data of the Japanese statute law. The ontology is inspired from PROV-O and SIOC ontologies. The LOD dataset based on the proposed ontology enables wide variety of analyses on the law history data by simple SPARQL queries. The analyses include simple search, visualization, temporal analysis, data mining, etc. This paper displays parts of the analyses which indicate several legislative facts behind changes of laws. The analyses demonstrate the proposed ontology and LOD dataset are useful for legal data analysis. The proposed ontology is comparable with ELI (European Legislation Identifier) which is designed for EU laws, this paper thus discusses the comparability and future directions of the proposed ontology.

Scopus

Publisher：Proceedings of the 33rd Pacific Asia Conference on Language, Information and Computation, PACLIC 2019  

We propose a method for finding and correct- ing misused Thai legal terms in Thai statu- tory sentences. Our method predicts legal terms using Random Forest classifiers, each of which is optimized for each set of similar legal terms. Each classifier utilizes outside- the-sentence features, namely, promulgation year, title keywords, and section keywords of statutes, in addition to words adjacent to the targeted legal term. Our experiment shows that our method outperformed not only a Ran- dom Forest method without the outside-the- sentence features, but also BERT (Bidirec- tional Encoder Representations from Trans- formers), a powerful language representation model, in overall accuracy.

Scopus

Language：English   Publishing type：Research paper (other academic)   Publisher：Springer  

DOI： 10.1007/978-3-030-36805-0_9

Scopus

Other Link： https://dblp.uni-trier.de/db/conf/ntcir/ntcir2019.html#OgawaSKT19

Publisher：CEUR Workshop Proceedings  

Entity search is a fundamental task in Linked Data (LD). The task is, given a keyword search query, to retrieve a set of entities in LD which are relevant to the query. The state-of-the-art approaches for entity search are based on information retrieval technologies such as TF-IDF vectorization and ranking models. This paper examines the approaches by applying a traditional evaluation metrics, recall@k, and shows ranking qualities still room left for improvements. In order to improve the ranking qualities, this paper explores possibilities of graph analytical methods. LD is regarded as a large graph, graph analytical approaches are therefore appropriate for this purpose. Since query-based graph analytical approaches fit to entity search tasks, this paper proposes a personalized PageRank-based re-ranking method, PPRSD (Personalized PageRank based Score Distribution), for retrieved results by the state-of-the-art. The experimental evaluation recognizes improvements but its results are not satisfactory, yet. For further improvements, this paper reports investigations about relationship between queries and entities in terms of path lengths on the graph, and discusses future directions for graph analytical approaches.

Scopus

Language：English  

Authorship：Lead author   Language：English  

Language：English  

DOI： 2-9

Language：English  

DOI： 334-339

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：Institute of Electrical and Electronics Engineers Inc.  

Analyzing and feeding back the results on real-world services are important missions in the Big Data era to realize smart city. However, analyzing real-world data is still challenging because of dirtiness of data and large variety of analytic requirements. To cope with the challenges, this paper proposes and develops an analytical toolbox for smart city applications. The analytical toolbox consists of three phases: preparation, analysis, and visualization. The preparation phase deals with the dirtiness of the data by including fundamental data cleansing techniques and data integration techniques. The analysis phase is responsible for ETL (extract, transform and load) process and analytical query processing from the next phase. The visualization phase deals with analytical requirements from users and visualization of analytical results. This paper showcases a real-world use case of the proposed analytical toolbox. The use case is now open in public with help of Fujisawa city, Japan, and this fact indicates that the proposed analytical toolbox is feasible for real-world data analysis and feeding back to citizens.

DOI： 10.1109/BigData.2017.8258429

Scopus

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：Institute of Electrical and Electronics Engineers Inc.  

Data integration is still laboursome task when integrating data are not consistently managed. Such inconsistency can happen easily in real-world situations, such as properties of objects are managed by a central organization and trajectories (or logs) of the objects are recorded by other peripheral organizations. This paper deals with a case of missing ordering information. Integrating property data and log data without ordering information causes duplicated results. In order to solve this problem, this paper proposes a join algorithm, called implicit order join, which discovers implicit ordering information from both property data and log data with help of partial true integrated results from human assistance. With the discovered ordering information, the implicit order join enables to integrate the property data and log data. In order to discover the implicit ordering information, ordering correlation between attribute sequences of property data and log data should be found from comprehensive examination of possible attribute sequence pairs. The potential number of sequence pairs is as high as factorial order of the number of attributes. Therefore, this paper develops a heuristic approach to prune unnecessary examinations based on ordering dependency between attribute sequences. Experimental evaluation in this paper indicates that implicit order join can reduce 77% labouring tasks for integration and the pruning method reduces the number of attribute sequences in orders of magnitude.

DOI： 10.1109/BigData.2017.8258474

Scopus

Language：English  

DOI： 10.3233/978-1-61499-935-5-161

Web of Science

Scopus

Publishing type：Research paper (international conference proceedings)   Publisher：Springer  

DOI： 10.1007/978-3-319-98809-2_10

Web of Science

Scopus

Other Link： https://dblp.uni-trier.de/db/conf/dexa/dexa2018-1.html#Komamizu18

Authorship：Lead author   Language：English  

DOI： 284-288

Authorship：Lead author   Language：English  

DOI： 35-42

Language：Japanese  

DOI： 138-145

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：Springer Verlag  

Ranking quality for keyword search over Linked Data (LD) is crucial when users look for entities from LD, since datasets in LD have complicated structures as well as much contents. This paper proposes a keyword search method, FORK, which ranks entities in LD by ObjectRank, a well-known link-structure analysis algorithm that can deal with different types of nodes and edges. The first attempt of applying ObjectRank to LD search reveals that ObjectRank with inappropriate settings gives worse ranking results than PageRank which is equivalent to ObjectRank with all the same authority transfer weights. Therefore, deriving appropriate authority transfer weights is the most important issue for encouraging ObjectRank in LD search. FORK involves a relevance feedback algorithm to modify the authority transfer weights according with users’ relevance judgements for ranking results. The experimental evaluation of ranking qualities using an entity search benchmark showcases the effectiveness of FORK, and it proves ObjectRank is more feasible raking method for LD search than PageRank and other comparative baselines including information retrieval techniques and graph analytic methods.

DOI： 10.1007/978-3-319-70145-5_5

Scopus

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：Knowledge Systems Institute Graduate School  

Analyzing behaviours of developers in different platforms (in particular, GitHub and Stack Overflow in this paper) can reveal interesting facts related to development activities. There are only few datasets for analysing crossplatform user behaviours, especially across GitHub and Stack Overflow. Users on GitHub and Stack Overflow are identifiable by equivalences of email addresses. In order to increase the number of identifiable users on these datasets, this paper retrieves potentially identifiable users between GitHub and Stack Overflow not relying only on email addresses. This paper employs a classification-based link prediction, which design the user identification problem as a link prediction problem on the bipartite graph consisting of users of GitHub and those of Stack Overflow. With the identification method, this paper generates a probabilistic dataset containing pairs of users with probabilities (or confidences). This paper, as well, publishes the identification tool in order to enable further data generation on appearing datasets of GitHub, Stack Overflow and others. The generated dataset and tool are highly helpful to accelerate researches on mining software repositories.

DOI： 10.18293/SEKE2017-109

Scopus

Authorship：Lead author   Language：English  

DOI： 10.1109/HPCC-SmartCity-DSS.2016.0192

Authorship：Lead author   Language：English  

DOI： 251-256

Authorship：Lead author   Language：English  

DOI： 10.1145/2938503.2938532

Authorship：Lead author   Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1108/IJWIS-03-2016-0014

Authorship：Lead author   Language：English  

DOI： 10.1145/2837185.2837230

Authorship：Lead author   Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1108/IJWIS-04-2015-0012

Authorship：Lead author   Language：English  

DOI： 10.1145/2684200.2684294

Authorship：Lead author   Language：English  

DOI： 10.1109/NBiS.2014.77

Authorship：Lead author   Language：English  

DOI： 10.1145/2628194.2628241

Authorship：Lead author   Language：English  

DOI： 10.1007/978-3-642-32597-7_40

Authorship：Lead author   Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1108/17440081211282865

Authorship：Lead author   Language：English  

DOI： 10.1145/2095536.2095544

Event date： 2018.9

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2018.9

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2018.9

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2018.9

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2018.9

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2018.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2018.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2017.4

Language：Japanese   Presentation type：Poster presentation  

Country：Japan  

Event date： 2017.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2017.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2014.9

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2012.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2012.2

Language：Japanese   Presentation type：Public lecture, seminar, tutorial, course, or other speech  

Country：Japan  

Event date： 2011.10

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2011.10

Language：English   Presentation type：Poster presentation  

Country：Australia  

Event date： 2011.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2011.2 - 2011.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2010.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2010.2 - 2010.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2009.9

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Event date： 2009.3

Language：Japanese   Presentation type：Oral presentation (general)  

Country：Japan  

Scopus

Scopus

Scopus

Scopus

Scopus

Scopus