Ko Fujimura

Geo topic model: joint modeling of user's activity area and interests for location recommendation

This paper proposes a method that analyzes the location log data of multiple users to recommend locations to be visited. The method uses our new topic model, called Geo Topic Model, that can jointly estimate both the users interests and activity area hosting the users home, office and other personal places. By explicitly modeling geographical features of locations and users, the users interests in other features of locations, which we call latent topics, can be inferred effectively. The topic interests estimated by our model 1) lead to high accuracy in predicting visit behavior as driven by personal interests, 2) make possible the generation of recommendations when the user is in an unfamiliar area (e.g. sightseeing), and 3) enable the recommender system to suggest an interpretable representation of the user profile that can be customized by the user. Experiments are conducted using real location logs of landmark and restaurant visits to evaluate the recommendation performance of the proposed method in terms of the accuracy of predicting visit selections. We also show that our model can estimate latent features of locations such as art, nature and atmosphere as latent topics, and describe each users preference based on them.

Topigraphy: visualization for large-scale tag clouds

This paper proposes a new method for displaying large-scale tag clouds. We use a topographical image that helps users to grasp the relationship among tags intuitively as a background to the tag clouds. We apply this interface to a blog navigation system and show that the proposed method enables users to find the desired tags easily even if the tag clouds are very large, 5,000 and above tags. Our approach is also effective for understanding the overall structure of a large amount of tagged documents.

Classical music for rock fans?: novel recommendations for expanding user interests

Most recommender algorithms produce types similar to those the active user has accessed before. This is because they measure user similarity only from the co-rating behaviors against items and compute recommendations by analyzing the items possessed by the users most similar to the active user. In this paper, we define item novelty as the smallest distance from the class the user accessed before to the class that includes target items over the taxonomy. Then, we try to accurately recommend highly novel items to the user. First, our method measures user similarity by employing items rated by users and a taxonomy of items. It can accurately identify many items that may suit the user. Second, it creates a graph whose nodes are users; weighted edges are set between users according to their similarity. It analyzes the user graph and extracts users that are related on the graph though the similarity between the active user and each of those users is not high. The users so extracted are likely to have highly novel items for the active user. An evaluation conducted on several datasets finds that our method accurately identifies items with higher novelty than previous methods.

Travel route recommendation using geotagged photos

We propose a travel route recommendation method that makes use of the photographers’ histories as held by social photo-sharing sites. Assuming that the collection of each photographer’s geotagged photos is a sequence of visited locations, photo-sharing sites are important sources for gathering the location histories of tourists. By following their location sequences, we can find representative and diverse travel routes that link key landmarks. Recommendations are performed by our photographer behavior model, which estimates the probability of a photographer visiting a landmark. We incorporate user preference and present location information into the probabilistic behavior model by combining topic models and Markov models. Based on the photographer behavior model, proposed route recommendation method outputs a set of personalized travel plans that match the user’s preference, present location, spare time and transportation means. We demonstrate the effectiveness of the proposed method using an actual large-scale geotag dataset held by Flickr in terms of the prediction accuracy of travel behavior.

The eigenrumor algorithm for calculating contributions in cyberspace communities

This paper describes a method for scoring the degree of contribution of each information object and each participant in a cyberspace community, e.g., knowledge management and product reviews, or other information sharing communities. Two types of actions, i.e., information provisioning and information evaluation, are common in such communities and are valuable in scoring each contribution. The EigenRumor algorithm, proposed here, calculates the contribution scores based on a link analysis approach by considering these actions as links from participants to information objects. The algorithm has similarities to Kleinbergs HITS algorithm in that both algorithms are based on the mutually reinforcing relationship of hubs and authorities but the EigenRumor model is not structured from page-to-page links but from participant-to-object links and is extended by the introduction of several new factors. The scores calculated by this algorithm can be used to identify “good” information and participants who contribute much to a community, which allows for the provisioning of incentives to such participants to promote their continuous contribution to the community.

Tweet classification by data compression

We propose a new method that uses data compression for classifying an unseen tweet as being related to an interesting topic or not. Our compression-based tweet classification method, called CTC, evaluates the compressibility of the tweet when given positive and negative examples. This enables our method to handle multilingual tweets in the same manner and to effectively utilize the word context of the tweet, which is extremely important information in the 140 character limit. Experiments with worldwide tweets assigned a single hashtag demonstrate that our method, which uses the Deflate algorithm (used in gzip) for empirical evaluations, achieved higher precision and recall rates than state-of-the-art online learning algorithms.

Topic structure mining using temporal co-occurrence

This paper proposes a topic structure mining method for document sets that include time stamps. Topic structure mining is a text mining method that uses the graph structure that represents the document pair similarities in the document set. This method yields not only topic extraction from documents and clustering of documents but also extracts the relationship between clusters and the meaning of each document in the cluster. Our method combines temporal co-occurrence with document similarity in constructing the graph structure. We also report evaluation results and the effectiveness of the proposed method.

Modeling multiple users' purchase over a single account for collaborative filtering

We propose a probabilistic topic model for enhancing recommender systems to handle multiple users that share a single account. In several web services, since multiple individuals may share one account (e.g. a family), individual preferences cannot be estimated from a simple perusal of the purchase history of the account, thus it is difficult to accurately recommend items to those who share an account. We tackle this problem by assuming latent users sharing an account and establish a model by extending Probabilistic Latent Semantic Analysis (PLSA). Experiments on real log datasets from online movie services and artificial datasets created from these real datasets by combining the purchase histories of two accounts demonstrate high prediction accuracy of users and higher recommendation accuracy than conventional methods.

A Voucher-Integrated Trading Model for C2B and C2C eCommerce System Development

This paper presents a trading model that aims to cover a wide variety of C2B and C2C business schemes. The proposed model uses a digital data entity called voucher and a voucher trading system (VTS) to construct a versatile trading framework. The trading framework handles the delivery/payment phases of trading transactions, and this makes it possible to construct various C2B and C2C e-commerce systems by simply combining the suitable matching logic implementations on our unified trading framework.

Trading Among Untrusted Partners Via Voucher Trading System

To provide highly usable electronic commerce systems at lower cost, it is important to utilize and co-ordinate the function-specific application service providers (ASPs) that are distributed throughout the Internet such as those providing matching, payment, and delivery services. In order to coordinate independent services that have not yet established trust among each other, this paper proposes to use electronic “vouchers” to link untrusted trading partners. A voucher is a digital representation of rights to claim goods or services and can be securely transferred between trading partners using the Voucher Trading System (VTS). This paper clarifies the basic functionalities that VTS should provide for coordinating untrusted trading partners.