Knowledge graph (KG)-enhanced recommender systems have been widely studied for alleviating data sparsity by incorporating rich relational semantics. However, existing KG-based models struggle to effectively model long-tail items due to the lack of reliable user-item supervision, which limits the exploitation of KG knowledge for recommendation. To address these challenges, we propose CHAKG, a cross-modal hyperedge alignment framework that compensates for missing supervision through efficient and selective interaction transfer. CHAKG projects both the KG and the user-item interaction graph into latent hyperedge spaces, enabling lightweight modeling of high-order semantic co-occurrence between interaction-rich items and long-tail items without expensive multi-hop propagation. Building on this representation, CHAKG introduces a consistency-guided interaction transfer mechanism, which selectively transfers reliable signals from interaction-rich items to long-tail items via a learnable denoising process. Furthermore, a cross-modal contrastive alignment objective enforces semantic consistency between knowledge and interaction views, improving robustness and generalization under sparse supervision. Extensive experiments on four benchmark datasets demonstrate that CHAKG achieves state-of-the-art accuracy and efficiency, particularly under cold-start and long-tail recommendation scenarios.

Reinforcement learning algorithms suffer from zero-shot generalization, where a learned policy is tested in new, out-of-distribution environments. One of the key reasons is the overfitting of the representation learning model. Previous studies have used data augmentation or contrastive learning to address this overfitting. Yet these methods often rely on task-specific augmentations or hand-picked negative samples, which are not robust to various tasks. We hypothesize that learning more abstract representations by ignoring pixel-level details allows the representation learning model to generalize better. Moreover, when an agent encounters a scene similar to one seen in the training environment but with slight variations in the background, it will ignore the differences and treat the scene as invariant, making the correct decision. Based on this idea, we introduce Dream with Abstractions (Dr. Abs), an algorithm that requires no custom augmentations or negative samples. First, we train a representation network with a joint embedding predictive architecture, where it learns an abstract representation. Next, we use the learned representation to regularize the latent states inside the reinforcement-learning agent. By ignoring pixel-level details, our method outperforms the baseline algorithm by approximately 22% on the widely used zero-shot generalization setup, thereby enhancing its generalization ability. Furthermore, Dr. Abs can be used in conjunction with other RL methods and can work in tandem with techniques such as data augmentation.

Causality stands as an essential relation for elucidating the reasoning behind given contents. However, current causality knowledge graphs fall short in effectively illustrating the inner logic in a specific domain, i.e. finance. To generate such a functional knowledge graph, we propose the multi-faceted approach encompassing causality detection module, entity linking module, and causality alignment module to automatically construct FinCaKG-Onto with the guidance of expert financial ontology - FIBO. In this paper, we outline the resources and methodology employed for FinCaKG-Onto construction, present the schema of FinCaKG-Onto, and share the final knowledge graph FinCaKG-Onto. Through various user scenarios, we demonstrate that FinCaKG-Onto not only captures nuanced domain expertise but also explicitly unveils the causal logic for any anchor terms. To facilitate your convenience of future use, a check table is conducted as well to showcase the quality of FinCaKG-Onto. The related resources are available in the webpage< https://www.ai.iee.e.titech.ac.jp/FinCaKG-Onto/ >.

Knowledge graphs (KGs) have been shown to be effective in improving recommendation quality by introducing rich item properties as auxiliary information. The success of current KG-based recommender systems (RSs) lies in the capability of modeling high-quality item representations. This is achieved by identifying significant properties for items and exploring the intrinsic correlation between items on the KG. However, since current KG-based works only focus on learning user implicit knowledge from KGs through items, the limited user-item interaction behavior is still an obstacle to learning high-quality user representations. Furthermore, irrelevant connections in the KG may lead to erroneous messaging during the process of high-order graph feature learning of users and items. This could subsequently result in the inaccurate recommendation of items to users. To overcome above limitations, we propose a Review-enhanced Contrastive Learning on KGs (RCLKG) model for high-quality recommendation. We first construct a review-enhanced KG by exploring user explicit preferences in reviews with the extracted review entities. Then, we design a review-aware self-augmentation mechanism that seamlessly integrates explicit review knowledge with item-aligned KGs to discard irrelevant neighbor nodes of users and items. Furthermore, we develop a global-level graph aggregation schema with a refined constraint on the merged denoising KG to further optimize the denoising KG generation by considering high-order connections with less erroneous messaging. Finally, experimental results on the rating prediction and the click-through rate prediction (CTR) tasks with three real-word datasets demonstrate the superiority of our proposed RCLKG model in comparison with the state-of-the-art baselines. • A novel review-enhanced contrastive learning model on KGs called RCLKG is proposed. • Review knowledge is injected into the knowledge graph as explicit knowledge of users. • Review-enhanced self-augmentation mechanism filters out irrelevant nodes in the KG. • Global-level graph encoder on merged denoising KG refines denoising KG generation. • Extensive experiments demonstrate the superiority of RCLKG.

Instrumental Variable (IV) provides a source of treatment randomization that is conditionally independent of the outcomes, responding to the challenges of counterfactual and confounding biases. In finance, IV construction typically relies on pre-designed synthetic IVs, with effectiveness measured by specific algorithms. This classic paradigm cannot be generalized to address broader issues that require more and specific IVs. Therefore, we propose an expertise-driven model (ETE-FinCa) to optimize the source of expertise, instantiate IVs by the expertise concept, and interpret the cause-effect relationship by integrating concept with real economic data. The results show that the feature selection based on causal knowledge graphs improves the classification performance than others, with up to a 11.7% increase in accuracy and a 23.0% increase in F1-score. Furthermore, the high-quality IVs we defined can identify causal relationships between the treatment and outcome variables in the Two-Stage Least Squares Regression model with statistical significance.

In the challenging context of Knowledge Graph (KG) construction from text, traditional approaches often rely on Open Information Extraction (OpenIE) pipelines. However, they are prone to generating many incorrect triplets. While domain specific Named Entity Recognition (NER) is commonly used to enhance the results, it compromises the domain independence and misses crucial triplets. To address these limitations, we introduce G-T2KG , a novel pipeline for KG construction that aims to preserve the domain independence while reducing incorrect triplets, thus offering a cost-effective solution without the need for domain-specific adaptations. Our pipeline utilizes state-of-the-art OpenIE combined with both a noun phrase-based cleaning and a LLMs based validation. It is evaluated using gold standards in two distinct domains (i.e., computer science and music) that we have constructed in the context of this study. On computer science corpus, the experimental results demonstrate a higher recall as compared to state-of-the-art approaches, and a higher precision notably increased by the integration of LLMs. Experiments on the music corpus show good performance, underscoring the versatility and effectiveness of G-T2KG in domain-independent KG construction.

Abstract Knowledge Graph Embedding (KGE) translates entities and relations of knowledge graphs (KGs) into a low-dimensional vector space, enabling an efficient way of predicting missing facts. Generally, KGE models are trained with positive and negative examples, discriminating positives against negatives. Nevertheless, KGs contain only positive facts; KGE training requires generating negatives from non-observed ones in KGs, referred to as negative sampling. Since KGE models are sensitive to inputs, negative sampling becomes crucial, and the quality of the negatives becomes critical in KGE training. Generative adversarial networks (GAN) and self-adversarial methods have recently been utilized in negative sampling to address the vanishing gradients observed with early negative sampling methods. However, they introduce the problem of false negatives with high probability. In this paper, we extend the idea of reducing false negatives by adopting a Tucker decomposition representation, i.e., TuckerDNCaching, to enhance the semantic soundness of latent relations among entities by introducing a relation feature space. TuckerDNCaching ensures the quality of generated negative samples, and the experimental results reflect that our proposed negative sampling method outperforms the existing state-of-the-art negative sampling methods.

Abstract Recent success in scaling deep reinforcement algorithms (DRL) to complex problems has been driven by well-designed extrinsic rewards, which limits their applicability to many real-world tasks where rewards are naturally extremely sparse. One solution to this problem is to introduce human guidance to drive the agent’s learning. Although low-level demonstrations is a promising approach, it was shown that such guidance may be difficult for experts to demonstrate since some tasks require a large amount of high-quality demonstrations. In this work, we explore human guidance in the form of high-level preferences between sub-goals, leading to drastic reductions in both human effort and cost of exploration. We design a novel hierarchical reinforcement learning method that introduces non-expert human preferences at the high-level, and curiosity to drastically speed up the convergence of subpolicies to reach any sub-goals. We further propose a strategy based on curiosity to automatically discover sub-goals. We evaluate the proposed method on 2D navigation tasks, robotic control tasks, and image-based video games (Atari 2600), which have high-dimensional observations, sparse rewards, and complex state dynamics. The experimental results show that the proposed method can learn significantly faster than traditional hierarchical RL methods and drastically reduces the amount of human effort required over standard imitation learning approaches.

Word embeddings or word vectors have become fundamental in language processing techniques, especially deep learning approaches. Although many languages have compound words (e.g., “robot arm” and “maple leaf”), such words have not received much attention from researchers. Most research on compound word embeddings considered only two-word compounds; there has been little detailed analysis on the learning representations of arbitrary-length compound words. This paper discusses the necessity for learning-based approaches for estimating the distributed representations of compound words instead of a simple average of the representations of constituents. An evaluation of two downstream tasks confirms the effectiveness of compositional models in encoding useful information into vector spaces. The experimental results suggest that complex architectures such as long short-term memory, gated recurrent units, and transformers learn better representations for long entities, whereas simpler models such as recurrent neural networks are more applicable for downstream tasks where there are only short compounds (two or three words in length), as in the noun compound interpretation task.

Abstract Deep reinforcement learning methods have achieved significant successes in complex decision-making problems. In fact, they traditionally rely on well-designed extrinsic rewards, which limits their applicability to many real-world tasks where rewards are naturally sparse. While cloning behaviors provided by an expert is a promising approach to the exploration problem, learning from a fixed set of demonstrations may be impracticable due to lack of state coverage or distribution mismatch—when the learner’s goal deviates from the demonstrated behaviors. Besides, we are interested in learning how to reach a wide range of goals from the same set of demonstrations. In this work we propose a novel goal-conditioned method that leverages very small sets of goal-driven demonstrations to massively accelerate the learning process. Crucially, we introduce the concept of active goal-driven demonstrations to query the demonstrator only in hard-to-learn and uncertain regions of the state space. We further present a strategy for prioritizing sampling of goals where the disagreement between the expert and the policy is maximized. We evaluate our method on a variety of benchmark environments from the Mujoco domain. Experimental results show that our method outperforms prior imitation learning approaches in most of the tasks in terms of exploration efficiency and average scores.

Abstract Deep reinforcement learning (DRL) algorithms rely on carefully designed environment rewards that are extrinsic to the agent. However, in many real-world scenarios rewards are sparse or delayed, motivating the need for discovering efficient exploration strategies. While intrinsically motivated agents hold promise of better local exploration, solving problems that require coordinated decisions over long-time horizons remains an open problem. We postulate that to discover such strategies, a DRL agent should be able to combine local and high-level exploration behaviors. To this end, we introduce the concept of fast and slow curiosity that aims to incentivize long-time horizon exploration. Our method decomposes the curiosity bonus into a fast reward that deals with local exploration and a slow reward that encourages global exploration. We formulate this bonus as the error in an agent’s ability to reconstruct the observations given their contexts. We further propose to dynamically weight local and high-level strategies by measuring state diversity. We evaluate our method on a variety of benchmark environments, including Minigrid, Super Mario Bros, and Atari games. Experimental results show that our agent outperforms prior approaches in most tasks in terms of exploration efficiency and mean scores.

A knowledge graph becomes a central data hub in the enterprise and the research communities. Nevertheless, the development of knowledge graphs is challenging due to the insufficient functionalities of knowledge graph management platforms. In this paper, we develop a knowledge graph management platform (UWKGM). This platform enables users to integrate arbitrary functionalities as RESTful API services in order to facilitate the knowledge graph development process. In the demonstration, we highlight the main features of UWKGM and its use cases on knowledge graph management tasks.

Deep reinforcement learning (DRL) methods traditionally struggle with tasks where environment rewards are sparse or delayed, which entails that exploration remains one of the key challenges of DRL. Instead of solely relying on extrinsic rewards, many state-of-the-art methods use intrinsic curiosity as exploration signal. While they hold promise of better local exploration, discovering global exploration strategies is beyond the reach of current methods. We propose a novel end-to-end intrinsic reward formulation that introduces high-level exploration in reinforcement learning. Our curiosity signal is driven by a fast reward that deals with local exploration and a slow reward that incentivizes long-time horizon exploration strategies. We formulate curiosity as the error in an agent’s ability to reconstruct the observations given their contexts. Experimental results show that this high-level exploration enables our agents to outperform prior work in several Atari games.

Abstract Reinforcement learning methods rely on rewards provided by the environment that are extrinsic to the agent. However, many real-world scenarios involve sparse or delayed rewards. In such cases, the agent can develop its own intrinsic reward function called curiosity to enable the agent to explore its environment in the quest of new skills. We propose a novel end-to-end curiosity mechanism for deep reinforcement learning methods, that allows an agent to gradually acquire new skills. Our method scales to high-dimensional problems, avoids the need of directly predicting the future, and, can perform in sequential decision scenarios. We formulate the curiosity as the ability of the agent to predict its own knowledge about the task. We base the prediction on the idea of skill learning to incentivize the discovery of new skills, and guide exploration towards promising solutions. To further improve data efficiency and generalization of the agent, we propose to learn a latent representation of the skills. We present a variety of sparse reward tasks in MiniGrid, MuJoCo, and Atari games. We compare the performance of an augmented agent that uses our curiosity reward to state-of-the-art learners. Experimental evaluation exhibits higher performance compared to reinforcement learning models that only learn by maximizing extrinsic rewards.

Knowledge graphs are useful for many AI tasks but often have missing facts. To populate the graphs, knowledge graph embedding models have been developed. TransE is one of such models and the first translation-based method. TransE is well known because the principle of TransE can effectively capture the rules of a knowledge graph although it seems very simple. However, TransE has problems with its regularization and an unchangeable ratio of negative sampling. In this paper, we generalize TransE to solve these problems by proposing knowledge graph embedding on a Lie group (KGLG) and the Weighted Negative Part (WNP) method for the objective function of translation-based models. KGLG is the novel translation-based method which embeds entities and relations of a knowledge graph on any Lie group. It allows us not to employ regularization during training of the model if we choose a compact lie group for the embedding space. The WNP method is for changing the ratio of negative sampling, which enhances translation-based models. Our approach outperforms other state-of-the-art approaches such as TransE, DistMult, and ComplEx on a standard link prediction task. We show that TorusE, KGLG on a torus, is scalable to large-size knowledge graphs and faster than the original TransE.

Power systems are becoming more complex, which increases instability issues and outage risks. The development of smart grids could help manage such complex systems. One important pillar in smart grids is big data analytics. In this poster paper, we discuss where and how machine learning could contribute to more efficient asset management. We also identify challenges that stand in the way of the widespread use of big data analytics in smart grids. While the nature of data, as well as data and asset management systems themselves make the use of big data challenging, data analytics could improve the reliability of power supply by providing the functions of detection, prediction, and selection.

Nowadays, modular domain ontology, where each module represents a subdomain of the ontology domain, facilitates the reuse of information and provides users with domain-specific knowledge. In this paper, we focus on modular taxonomy learning from text, where each module collects terms with the same topic insights, and in parallel we manage to discover hypernym and 'related' relations among those collected terms.However, it is difficult to automatically fit terms into modules and discover relations.We propose to employ twice trainedLDAto partition termsof each subdomain, and relate subdomains into modules of ontology. Meanwhile, we apply core concept replacement and subdomain knowledge supplementation as supportive information embedding technique over the corpus. This shows that the twice trained LDA strategy can effectively identify topic-relevant terms into subdomains, with nearly two-fold precision comparing to that of normal LDA training. The combination of core concept replacement and subdomain knowledge supplementation contributes to significant improvements in modular taxonomy learning.

This study analyzes the value conflict of a paper on fan fiction writing that used online fan fiction novels as a source to extract and filter sexual expressions from text. The boundaries of public and private information are ambiguous because users are not always aware of or have agreed to the fact that their content is to be used openly. The case was complicated by the fact that the use of these data by researchers violated an unconsciously infringed upon right of a vulnerable community with a weak legal position. This paper describes the debate on this topic among researchers from engineering and humanities fields on whether the purpose of the research was ethically acceptable; how the systems can be embedded in ethical values; and what ethical, legal, social, and educational lessons are appropriate for governance of artificial intelligence (AI). Our analysis aimed not only to clarify the abstract concept of privacy but also to make changes to the submission guidelines for authors. We hope that our analysis contributes to the governance of ethical AIs and AI ethics on handling sensitive aspects of online activities.

Knowledge graphs are useful for many artificial intelligence (AI) tasks. However, knowledge graphs often have missing facts. To populate the graphs, knowledge graph embedding models have been developed. Knowledge graph embedding models map entities and relations in a knowledge graph to a vector space and predict unknown triples by scoring candidate triples. TransE is the first translation-based method and it is well known because of its simplicity and efficiency for knowledge graph completion. It employs the principle that the differences between entity embeddings represent their relations. The principle seems very simple, but it can effectively capture the rules of a knowledge graph. However, TransE has a problem with its regularization. TransE forces entity embeddings to be on a sphere in the embedding vector space. This regularization warps the embeddings and makes it difficult for them to fulfill the abovementioned principle. The regularization also affects adversely the accuracies of the link predictions. On the other hand, regularization is important because entity embeddings diverge by negative sampling without it. This paper proposes a novel embedding model, TorusE, to solve the regularization problem. The principle of TransE can be defined on any Lie group. A torus, which is one of the compact Lie groups, can be chosen for the embedding space to avoid regularization. To the best of our knowledge, TorusE is the first model that embeds objects on other than a real or complex vector space, and this paper is the first to formally discuss the problem of regularization of TransE. Our approach outperforms other state-of-the-art approaches such as TransE, DistMult and ComplEx on a standard link prediction task. We show that TorusE is scalable to large-size knowledge graphs and is faster than the original TransE.

Word embedding is the state-of-the-art representation to capture semantic information of terms. It benefits a wide range of natural language processing and related applications, not only in general fields of artificial intelligence but also in bioinformatics. Although recent efforts of using word embedding to represent medical concepts have provided remarkable analyses, many essential problems remain unsolved. Examples include representation of complex concepts (i.e., formed by multiple tokens), leveraging of a large corpus to maximize the trainable concepts, and downstream analyses on a biomedical-related dataset. Our study focused on training effective representations for biomedical concepts including complex ones. We used an efficient technique to index all possible concepts of UMLS thesaurus (Unified Medical Language System) in a huge corpus of 15,4 billion tokens. By this way, we can obtain the vector representations for more than 650,000 concepts, the largest ever reported resource to date. Furthermore, evaluations of trained vectors on retrieval task show superior performance compared to recent studies.

The increased penetration of distributed and volatile renewable generation requires the demand-side to be actively involved in energy balancing operations. This paper proposes a solution in which big data and machine learning methods are employed to enhance the capabilities of a Virtual Power Plant to participate and intelligently bid into a demand response energy market. The energy market being investigated consists of the frequency containment reserve market. First, we define the core decision-making required to overcome the uncertainties in the frequency containment reserve market participation for a Virtual Power Plant. Then, we focus on forecasting the frequency containment reserve prices for the day-ahead. We analyze the price data, and identify and collect the relevant features for the prediction of the prices. In addition, we select several regression analysis methods to be utilized for the prediction. Finally, we evaluate the performance of the implemented methods by executing several experiments, and compare the the performance with the performance of a state of the art autoregression method.

Due to the rise of artificial intelligence (AI) technology, discussions are progressing on how robots could replace human labor. Conventional surveys have suggested that human labor is expected to gradually be replaced as tasks become automated. We conducted a survey at the world's first robot hotel recently opened in Japan - called a Henn-na hotel (“strange/change hotel”) in Japanese - which already uses robots for most of the work. We discovered that human labor is divided into small tasks, and that robot actions affect human emotional control. However, the hotel not only divides human work but also reconstructs it from tasks. Moreover, the purpose of reconstruction is not simply for replacement of works. Such modification of task is often observed taking place in humansystem interactions. It is an extremely creative process of labor emerging in this area.

The goal of linked data instance matching is to detect all instances that co-refer to the same objects in two linked data repositories, the source and the target. Since the amount of linked data is rapidly growing, it is important to automate this task. However, the difference between the schemata of source and target repositories remains a challenging barrier. This barrier reduces the portability, accuracy, and scalability of many proposed approaches. The authors present automatic schema-independent interlinking (ASL), which is a schema-independent system that performs instance matching on repositories with different schemata, without prior knowledge about the schemata. The key improvements of ASL compared to previous systems are the detection of useful attribute pairs for comparing instances, an attribute-driven token-based blocking scheme, and an effective modification of existing string similarities. To verify the performance of ASL, the authors conducted experiments on a large dataset containing 246 subsets with different schemata. The results show that ASL obtains high accuracy and significantly improves the quality of discovered coreferences against recently proposed complex systems.

Making fast driving decisions at intersections is a challenging problem for improving safety of autonomous vehicles. Furthermore, representing sensor data in a machine understandable format is essential to enable vehicles to understand traffic situations. Ontologies are used to represent knowledge of sensor data for autonomous vehicles to aware traffic situations. In this paper, we introduce a fast decision making system, which utilizes only related part of the ontology-based knowledge base to make decisions at intersections. The decision making system performs real-time reasoning using traffic regulations and a part of the map information from the knowledge base.

Many Advanced Driver Assistance Systems (ADAS) have been developed to improve car safety. However, it is still a challenging problem to make autonomous vehicles to drive safely on urban streets such as uncontrolled intersections (without traffic lights) and narrow roads. In this paper, we introduce a decision making system that can assist autonomous vehicles at uncontrolled intersections and narrow roads. We constructed a machine understandable ontology-based Knowledge Base, which contains maps and traffic regulations. The system makes decisions in comply with traffic regulations such as Right-Of-Way rules when it receives a collision warning signal. The decisions are sent to a path planning system to change the route or stop to avoid collisions.

Multi-class classification is mandatory for real world problems and one of promising techniques for multi-class classification is Error Correcting Output Code. We propose a method for constructing the Error Correcting Output Code to obtain the suitable combination of positive and negative classes encoded to represent binary classifiers. The minimum weight perfect matching algorithm is applied to find the optimal pairs of subset of classes by using the generalization performance as a weighting criterion. Based on our method, each subset of classes with positive and negative labels is appropriately combined for learning the binary classifiers. Experimental results show that our technique gives significantly higher performance compared to traditional methods including One-Versus-AU, the dense random code, and the sparse random code. Moreover, our method requires significantly smaller number of binary classifiers while maintaining accuracy compared to One-Versus-One.

A supergraph containment search is to retrieve the data graphs contained by a query graph. In this paper, we study the problem of efficiently retrieving all data graphs approximately contained by a query graph, namely similarity search on supergraph containment. We propose a novel and efficient index to boost the efficiency of query processing. We have studied the query processing cost and propose two index construction strategies aimed at optimizing the performance of different types of data graphs: top-down strategy and bottom-up strategy. Moreover, a novel indexing technique is proposed by effectively merging the indexes of individual data graphs; this not only reduces the index size but also further reduces the query processing time. We conduct extensive experiments on real data sets to demonstrate the efficiency and the effectiveness of our techniques.

In this study, we proposed the use of grant application data to acquire knowledge of the relationships among scientific research subjects. We modeled grant application data to construct a method of capturing the relationships among research subjects, then conducted experiments using actual grant application data. The results indicated that our method successfully elucidated the relationships among research subjects.

As information technology progress, we are able to obtain much information about the advanced research of others. As a result, researchers and research managers need to track the current research trends amid the information flood. In order to support these efforts to gather knowledge of current research, we propose a research trend mining method. The method utilizes an author-topic model for establishing the relationships between authors, topics, and papers by probabilities, and interactively visualizes the relationships using self-organizing maps. We implemented a research area mapping system and validated it with a case study. In addition, we conducted experiments to show the performance of our system. The experimental results indicate that this system can induce the appropriate relationships for finding research trends.

This paper reports a bibliometric analysis of evolving co-author networks. Using 5,009 articles covering the years 1993 to 2005 from Transactions D. (Information Systems) of the Institute of Electronics Information and Communication Engineers (IEICE), we attempt to compare the network characteristics for each year, the co-author network characteristics for the entire time span, and the four major components of the entire data set. Finally, we analyze each of these in contrast to extant co-authorship network data and find that the pattern of co-authorship within Information Systems does not change significantly over this time period.

Since research trends can change dynamically, researchers have to keep up with these new trends and undertake new research topics. Therefore, research communities for new research domains are important. In this paper, we propose a method to discover research communities. The key features of our method are a network model of papers and a word assignment technique for the communities obtained. We show our system based on the proposed method and discuss our system through case studies and experiments

Research communities are very important for researchers undertaking new research topics. In this paper, we propose a community mining system using bibliography data in order to find communities of researchers. The basic concept of our study is to provide interactive visualization of both local and global research communities. We implement this concept using actual bibliography data and present a case study using the proposed system.