Search

Nowadays, domain ontologies are widely used as background knowledge bases. However, end users of ontology-based question answering (QA) systems are unaware of major concepts of ontology or the structure of domain ontology schema. Thus, it has been essential to provide an efficient method to reduce this gap. Namely, the critical issue for ontology-based QA systems is how to generate a SPARQL query from a user’s natural language query (NLQ). Therefore, we proposed a method to generate SPARQL queries from Korean natural language queries. When an input query comes in, we split it into a set of tokens and map each token to certain resources in the ontology. Subsequently, a graph generation process creates multiple “query graphs” by arranging the resources and identifying relationships between them. To identify relations between resources, we applied a path search algorithm based on the structure of domain ontology schema. We score query graphs by measuring the degree to which the graph reflected the general user’s intent, and the best-rated query graph is converted into a SPARQL query. We implemented a prototype system to evaluate the proposed method for the music domain ontology and conclude that our query conversion process can convert Korean natural language queries into semantically equivalent SPARQL queries. We anticipate that, after appropriate modification, the process can be applied to other languages.

In many areas, vast amounts of information are rapidly accumulating in the form of ontology-based knowledge graphs, and the use of information in these forms of knowledge graphs is becoming increasingly important. This study proposes a novel method for efficiently learning frequent subgraphs (i.e., knowledge) from ontology-based graph data. An ontology-based large-scale graph is decomposed into small unit subgraphs, which are used as the unit to calculate the frequency of the subgraph. The frequent subgraphs are extracted through candidate generation and chunking processes. To verify the usefulness of the extracted frequent subgraphs, the methodology was applied to movie rating prediction. Using the frequent subgraphs as user profiles, the graph similarity between the rating graph and new item graph was calculated to predict the rating. The MovieLens dataset was used for the experiment, and a comparison showed that the proposed method outperformed other widely used recommendation methods. This study is meaningful in that it proposed an efficient method for extracting frequent subgraphs while maintaining semantic information and considering scalability in large-scale graphs. Furthermore, the proposed method can provide results that include semantic information to serve as a logical basis for rating prediction or recommendation, which existing methods are unable to provide.

Research and development projects in railroad domain are different from those in other domains in terms of their close relationship with laws. Some cases are reported that new technologies from R&D projects could not be industrialized because of relevant laws restricting them. This problem comes from the fact that researchers don’t know exactly what laws can affect the result of R&D projects. To deal with this problem, we suggest a model for law retrieval system that can be used by researchers of railroad R&D projects to find related legislation. Input of this system is a research plan describing the main contents of projects. After laws related to the R&D project is provided with their rankings, which are assigned by scores we developed. A ranking of a law means its order of priority to be checked. By using this system, researchers can search the laws that may affect R&D projects throughout all the stages of project cycle. So, using our system model, researchers can get a list of laws to be considered before the project they participate ends. As a result, they can adjust their project direction by checking the law list, avoiding their elaborate projects being useless.

Research and development projects have close relationship with laws. In some cases, new technologies resulted from R&D projects can't be used because some statutes restrict them. The reason of this problem is that researchers don't know exactly which laws can affect their R&D projects. To solve the issue, we suggest a model for law retrieval system that can be used by researchers of R&D projects to find related statutes. Input of this model is a query document that describes the main contents of a project. By using ontology, legal terms are extracted from the document and statutes defining them are retrieved as a set of related laws. After this searching process, statutes are provided to researchers with their ranks, which are assigned using relevance scores we developed. By using this model, we can make a system for researchers to search a list of statutes that may affect R&D projects, and finally, they can adjust their project's direction by checking the list, preventing their works from being useless.