Hee-Geun Yoon

An automatic translation of tags for multimedia contents using folksonomy networks

This paper proposes a novel method to translate tags attached to multimedia contents for cross-language retrieval. The main issue in this problem is the sense disambiguation of tags given with few textual contexts. In order to solve this problem, the proposed method represents both tags and its translation candidates as networks of co-occurring tags since a network allows richer expression of contexts than other expressions such as co-occurrence vectors. The method translates a tag by selecting the optimal one from possible candidates based on a network similarity even when neither the textual contexts nor sophisticated language resources are available. The experiments on the MIR Flickr-2008 test set show that the proposed method achieves 90.44% accuracy in translating tags from English into German, which is significantly higher than the baseline methods of a frequency based translation and a co-occurrence-based translation.

A Translation-Based Knowledge Graph Embedding Preserving Logical Property of Relations

This paper proposes a novel translation-based knowledge graph embedding that preserves the logical properties of relations such as transitivity and symmetricity. The embedding space generated by existing translation-based embeddings do not represent transitive and symmetric relations precisely, because they ignore the role of entities in triples. Thus, we introduce a role-specific projection which maps an entity to distinct vectors according to its role in a triple. That is, a head entity is projected onto an embedding space by a head projection operator, and a tail entity is projected by a tail projection operator. This idea is applied to TransE, TransR, and TransD to produce lppTransE, lppTransR, and lppTransD, respectively. According to the experimental results on link prediction and triple classification, the proposed logical property preserving embeddings show the state-of-the-art performance at both tasks. These results prove that it is critical to preserve logical properties of relations while embedding knowledge graphs, and the proposed method does it effectively.

Using MCRDR based Agile approach for expert system development

Various expert system development approaches were proposed but most of them cannot deal with two problems: the difficulty of analysis and maintenance. Rather than to spend time waiting any longer, it is better to find an alternative solution from other research fields. In computer software development area, researchers have been suffering from the difficulty of maintenance and analysis, just as the researchers in the expert system development field. To solve this issue, researchers in the software used both agile software development and business rules approach: agile software development is for overcoming the the difficulty of analysis, and business rules approach is for reducing issues in the maintenance. There is a big opportunity that those two approaches can also be solve the two issues in the expert system development field. The paper describes requirements of the approach based on agile software development and the business rules approach. As a result, we consider and specify why the Multiple Classification Ripple Down Rules is the novel approach for the expert system development.

Ontology Population from Unstructured and Semi-structured Texts

Legacy information search systems have limitation that it does not consider semantic information but just lexical information such as keywords. A semantic web is expected to solve such limitation of present systems. In constructing semantic web, an ontology is believed to be a must. However, the ontology construction is very difficult. It requires great human efforts, since the creation of individuals is a time consuming task. Thus, there is a potential need for automatic or semiautomatic ontology population system, which greatly alleviates the human efforts. This paper proposes a method for an ontology population, in which the population is processed by computing the overlap between instances and concepts. This method is very simple but shows high performance.

Experience Search: Accessing the Emergent Knowledge from Annotated Blog Postings

Emergent knowledge does not come from a particular document or a particular knowledge source, but comes from a collection of documents or knowledge sources. This paper proposes a system which combines the social web contents and the semantic web technology to process the emergent knowledge from the blogosphere. The proposed system regards blog postings as experiences of people on particular topics. By annotating postings in the selected domains with ontology vocabularies, the system collects experiences from various people into an ontology about people and experiences. The system processes this ontology with semantic rules to find the emergent knowledge. Users can access previously unavailable facts, concepts and trends which are emerging from system.

Query Expansion based on Word Sense Community

In order to assist users who are in the process of executing a search, a query expansion method suggests keywords that are related to an input query. Recently, several studies have suggested keywords that are identified by finding domains using a clustering method over the documents that are retrieved. However, the clustering method is not relevant when presenting various domains because the number of clusters should be fixed. This paper proposes a method that suggests keywords by finding various domains related to the input queries by using a community detection algorithm. The proposed method extracts words from the top-30 documents of those that are retrieved and builds communities according to the word graph. Then, keywords representing each community are derived, and the represented keywords are used for the query expansion method. In order to evaluate the proposed method, we compared our results to those of two baseline searches performed by the Google search engine and keyword recommendation using TF-IDF in the search results. The results of the evaluation indicate that the proposed method outperforms the baseline with respect to diversity.

Ontology Kernel A Convolution Kernel for Ontology Alignment

Every ontology entity such as a concept or a property has its own structural information represented as a graph due to the relations with other entities. Therefore, it is important to consider not only its lexical similarity but also structural similarity in ontology alignment. This paper proposes ontology kernel that computes both types of similarities simultaneously. The idea of this kernel is to measure the structural similarity of ontology entities by mapping their entity graphs into the space spanned by entity random walks. The graph of an entity in the kernel expresses all relations with other entities. Thus, the ontology kernel can compare the similarity between entities no matter how complex the entities are and no matter how many kinds of relations they possess. A series of experiments with the standard data sets prove the generality and the superiority of the ontology kernel in ontology alignment.

Determining Gender of Korean Names with Context

Machine translation systems have various problems although they have been developed continuously. Especially, in Korean-English translation system, zero pronoun problem is an important problem, since omitted subject or object Korean are must be restored in English. In order to solve this problem, various methods have been proposed. In this paper, we focus on the gender determination problem in Korean names as a first-step for solving a zero pronoun problem in Korean. Since this problem can be viewed as a binary classification problem, we adopt support vector machines which are well-known for solving binary classification. The bag-of-words model is used to represent a name with context as a vector and information entropy of words is adopted for selecting features. An evaluation of the proposed method shows about 86% of accuracy. This method achieves higher accuracy than baseline which determines the gender of a name by its majority and additionally resolves the limitation of memory based and statistical method which use only names.

A re-ranking model for accurate knowledge base completion with knowledge base schema and web statistic

Knowledge base completion aims to complete a knowledge base by filling up missing facts of the knowledge base. Neural knowledge base embeddings proposed to solve this task measure the plausibility of all candidate triples, and then select top-ranked triples by the plausibility as new facts for the knowledge base. The plausibility by neural embeddings allows true facts to be ranked at high positions, but not at top positions. This is because neural knowledge base embeddings are limited to using only the information within the knowledge base. Therefore, this paper proposes a re-ranking model for precise knowledge base completion. As a re-ranking model, a neural network which uses knowledge base schema and web statistic additionally is adopted. As a result, the proposed re-ranking model has an effect of using additional information for knowledge base completion. Thus, the candidate triples are first ranked by a neural knowledge base embedding, and then the result is re-ranked by the neural network. The experimental results show that the proposed re-ranking model improves the base neural embeddings up to 16% in Hits@1. This implies that the re-ranking model places true facts at top positions effectively.

Improving The Performance of Triple Generation Based on Distant Supervision By Using Semantic Similarity

The existing pattern-based triple generation systems based on distant supervision could be flawed by assumption of distant supervision. For resolving flaw from an excessive assumption, statistics information has been commonly used for measuring confidence of patterns in previous studies. In this study, we proposed a more accurate confidence measure based on semantic similarity between patterns and properties. Unsupervised learning method, word embedding and WordNet-based similarity measures were adopted for learning meaning of words and measuring semantic similarity. For resolving language discordance between patterns and properties, we adopted CCA for aligning bilingual word embedding models and a translation-based approach for a WordNet-based measure. The results of our experiments indicated that the accuracy of triples that are filtered by the semantic similarity-based confidence measure was 16% higher than that of the statistics-based approach. These results suggested that semantic similarity-based confidence measure is more effective than statistics-based approach for generating high quality triples.