著者

白井 嵩士 榊 剛史 鳥海 不二夫 篠田 孝祐 風間 一洋 野田 五十樹 沼尾 正行 栗原 聡 Shirai Takashi Sakaki Takeshi Toriumi Fujio Shinoda Kosuke Kazama Kazuhiro Noda Itsuki Numao Masayuki Kurihara Satoshi

雑誌

SIG-DOCMAS = SIG-DOCMAS

巻号頁・発行日

no.B102, 2012-03-11

---

Twitter is a famous social networking service and has received attention recently.Twitter user have increased rapidly, and many users exchange information. When 2011 Tohokuearthquake and tsunami happened, people were able to obtain information from social networkingservice. Though Twitter played the important role, one of the problem of Twitter, a false rumordiffusion, was pointed out. In this research, we focus on a false rumor diffusion. We propose ainformation diffusion model based on SIR model, and discuss how to prevent a false rumor diffusion.

著者

Nattee Cholwich Khamsemanan Nirattaya Theeramunkong Thanaruk Numao Masayuki

出版者

人工知能学会

雑誌

人工知能学会全国大会論文集 (ISSN:13479881)

巻号頁・発行日

vol.26, 2012

---

Positive and Unlabeled learning (PU learning) is a machine learning approach that focuses on generating a two-class classification model using only a set of positive examples, and a set of unlabeled examples. Various techniques have been proposed for PU learning. Most of the techniques try to detect a group of reliables negative examples from the given unlabeled examples. Then, a classification model can be incrementally built. In this paper, we propose a new technique for detecting the reliable negative examples based on the density of examples in the search space.

著者

Cabredo Rafael Inventado Paul Legaspi Roberto Numao Masayuki

出版者

人工知能学会

雑誌

人工知能学会全国大会論文集 (ISSN:13479881)

巻号頁・発行日

vol.26, 2012

---

Current music recommender systems only use basic information for recommending music to its listeners. These usually include artist, album, genre, tempo and other song information. Online recommender systems would include ratings and annotation tags by other people as well. We propose a recommender system that recommends music depending on how the listener wants to feel while listening to the music. The user-specific model we use is derived by analyzing brainwaves of the subject while he was actively listening to emotion-inducing music. The brainwaves are analyzed in order to derive the emotional state of the listener for different segments of the music. Using a motif discovery algorithm, we discover pairs of similar subsequences from the emotion data and find correlations with music and audio features from the song. Similar patterns are clustered and used for recommending music that invoke a similar emotional response from the listener.

著者

B. Mai Anh Legaspi Roberto Inventado Paul Cabredo Rafael Kurihara Satoshi Numao Masayuki

出版者

人工知能学会

雑誌

人工知能学会全国大会論文集 (ISSN:13479881)

巻号頁・発行日

vol.26, 2012

---

As more and more information find their way to the internet, people are able to do more at their own desk than ever before, all in the comfort of a private environment. But as more activities, especially learning, are able to be done through the personal desktop space, the question is then raised of whether or not one is really engaged and/or learning and not being distracted by other things that the internet offer. For this, we propose a model that will associate various sitting postures with a person's level of engagement and/or learning. Said model will know what kind of postures usually indicate a state of engagement to a person's work and learning, and which postures indicate a falling out from that state. We apply machine learning techniques to a database of silhouette images, captured using a Microsoft Kinect, in order to extrapolate patterns that would help link a user's postures to his learning state. Our model can be used to assist users regain learning postures and suggest for a change of activity if prolonged periods of non-learning are detected so that users will gain the most out of their time.