Briefly reviews International Conference on Weblogs and Social Media (ICWSM12) from my perspective.
The latter part written in Japanese, sorry for that.
This document provides an overview of clustering techniques including k-means clustering, expectation maximization algorithms, and spectral clustering. It discusses how k-means clustering works by initializing random cluster centers, assigning data points to the closest centers, and adjusting the centers iteratively. Expectation maximization is presented as a way to learn the parameters of a Gaussian mixture model to cluster data. Finally, applications of clustering like document clustering using mixture models are briefly described.
This document summarizes research on clustering blogs and discovering blog communities. It outlines the significance of clustering the huge and growing blogosphere. Both network-based and content-based clustering approaches are discussed, as well as hybrid approaches. Evaluation of approaches shows hybrid clustering using both network and accompanying content information leads to more coherent blog clusters and distinct communities compared to network-based information alone. The document concludes more work should consider temporal dynamics in blog clustering.
This document provides an overview of spectral clustering. It begins with a review of clustering and introduces the similarity graph and graph Laplacian. It then describes the spectral clustering algorithm and interpretations from the perspectives of graph cuts, random walks, and perturbation theory. Practical details like constructing the similarity graph, computing eigenvectors, choosing the number of clusters, and which graph Laplacian to use are also discussed. The document aims to explain the mathematical foundations and intuitions behind spectral clustering.
This document outlines the roadmap and agenda for a machine learning meetup covering clustering algorithms. The meetup will include sessions on k-means clustering, DBSCAN, hierarchical clustering, mean shift, spectral clustering and dimension reduction. Spectral clustering will be covered in two sessions focusing on the mathematical foundations and applications in computer vision. The meetup aims to provide an overview of machine learning techniques and their applications in domains such as business analytics, recommendation systems, natural language processing and the energy industry.
This document provides an overview of clustering techniques including k-means clustering, expectation maximization algorithms, and spectral clustering. It discusses how k-means clustering works by initializing random cluster centers, assigning data points to the closest centers, and adjusting the centers iteratively. Expectation maximization is presented as a way to learn the parameters of a Gaussian mixture model to cluster data. Finally, applications of clustering like document clustering using mixture models are briefly described.
This document summarizes research on clustering blogs and discovering blog communities. It outlines the significance of clustering the huge and growing blogosphere. Both network-based and content-based clustering approaches are discussed, as well as hybrid approaches. Evaluation of approaches shows hybrid clustering using both network and accompanying content information leads to more coherent blog clusters and distinct communities compared to network-based information alone. The document concludes more work should consider temporal dynamics in blog clustering.
This document provides an overview of spectral clustering. It begins with a review of clustering and introduces the similarity graph and graph Laplacian. It then describes the spectral clustering algorithm and interpretations from the perspectives of graph cuts, random walks, and perturbation theory. Practical details like constructing the similarity graph, computing eigenvectors, choosing the number of clusters, and which graph Laplacian to use are also discussed. The document aims to explain the mathematical foundations and intuitions behind spectral clustering.
This document outlines the roadmap and agenda for a machine learning meetup covering clustering algorithms. The meetup will include sessions on k-means clustering, DBSCAN, hierarchical clustering, mean shift, spectral clustering and dimension reduction. Spectral clustering will be covered in two sessions focusing on the mathematical foundations and applications in computer vision. The meetup aims to provide an overview of machine learning techniques and their applications in domains such as business analytics, recommendation systems, natural language processing and the energy industry.
□Author
Masaya Mori, Global Head of Rakuten Institute of Technology, Executive Officer, Rakuten Inc.
森正弥 楽天株式会社 執行役員 兼 楽天技術研究所代表
□Description
そもそもなぜ人工知能(AI)をビジネスで活用する必要があるのかの視点に基づいて、AI活用戦略について述べた講演の資料です。
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Akisato Kimura
Introducing the paper "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning" presented in ICML2016 (in Japanese).
Updated version of https://www.slideshare.net/akisatokimura/paper-reading-dropout-as-a-bayesian-approximation-representing-model-uncertainty-in-deep-learning
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Akisato Kimura
A stale version, please check https://www.slideshare.net/akisatokimura/paper-reading-dropout-as-a-bayesian-approximation-representing-model-uncertainty-in-deep-learning-166237519 for a new version.
Introducing the paper "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning" presented in ICML2016 (in Japanese).
□Author
Masaya Mori, Global Head of Rakuten Institute of Technology, Executive Officer, Rakuten Inc.
森正弥 楽天株式会社 執行役員 兼 楽天技術研究所代表
□Description
そもそもなぜ人工知能(AI)をビジネスで活用する必要があるのかの視点に基づいて、AI活用戦略について述べた講演の資料です。
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Akisato Kimura
Introducing the paper "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning" presented in ICML2016 (in Japanese).
Updated version of https://www.slideshare.net/akisatokimura/paper-reading-dropout-as-a-bayesian-approximation-representing-model-uncertainty-in-deep-learning
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Akisato Kimura
A stale version, please check https://www.slideshare.net/akisatokimura/paper-reading-dropout-as-a-bayesian-approximation-representing-model-uncertainty-in-deep-learning-166237519 for a new version.
Introducing the paper "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning" presented in ICML2016 (in Japanese).
NIPS2015 reading - Learning visual biases from human imaginationAkisato Kimura
1) The document discusses a paper on improving visual recognition systems by leveraging human visual biases and generating images from random features.
2) It describes estimating visual biases from human psychophysics experiments, then using those biases to reconstruct images from random features. The reconstructed images can then be used to train machine learning models.
3) The document outlines experiments showing that incorporating estimated human visual biases into machine learning models, such as SVMs, can help improve visual recognition performance compared to models trained without biases.
CVPR2015 reading "Global refinement of random forest"Akisato Kimura
- A method is presented for refining a pre-trained random forest by optimizing the leaf weights while keeping the tree structures fixed.
- This reformulates the random forest as a linear classification/regression problem where samples are represented by sparse indicator vectors.
- The optimization can be performed efficiently and the refined forest has comparable or better accuracy than the original forest, but with significantly fewer trees/nodes.
- Experiments on classification and regression datasets demonstrate the proposed method outperforms other random forest techniques while accelerating training and testing.
Computational models of human visual attention driven by auditory cuesAkisato Kimura
This document summarizes a presentation on computational models of human visual attention driven by auditory cues. It discusses how auditory information can modulate visual attention by selecting visual features that are synchronized with detected auditory events. The proposed model uses Bayesian surprise to detect transient events in visual and auditory streams separately, then correlates the two to select synchronized visual features. An evaluation of the model on video clips found it outperformed baseline models at predicting eye movements.
Brief description of the paper "Large-scale visual sentiment ontology and detectors using adjective noun pairs" presented in ACM Multimedia 2013 as a full paper.
4. What’s ICWSM?
International AAAI Conference on Weblogs
and Social Media
Annual conference, 6th for this year.
Seems to be a conference on Twitter & other
social media, few papers as to weblogs.
A lot of participants from companies and labs
about SNS, mass media, ads, and marketing.
A major cluster = sociologists,
a unique conference hosted by AAAI.
5. Symbolic panel discussions
I Want to (Net)work With You, But I Don't
Know What/Where/Who You Are
Panelists from Cisco, IBM, LinkedIn & Datahug
News Generation and Consumption Through
Social Media
Panelists from Storyful, Newswhip, Irish Times,
C-SPAN & Guardian
Machine learning accounts for a small portion.
6. Basic statistics
Only single track
Not high quality
as the rate indicates
Our presentation (can’t see any other JPN pres.)
Attendees: over 330 in advanced registration (x3 of papers),
half of them from USA, only 5 from Japan.
7. General overview
Computer science << sociology
Data collecting, analyses & discussions
> results > performance > technical novelty
Most oral presentations with high quality
Especially in terms of analysis and discussions.
Don’t mind theoretical soundness and novelty.
2 giants: Twitter & Facebook
But, we should not rely only on the giants.
The direction includes cross platform analysis.
8. Interesting events & efforts
Town hall meeting
Discussing future directions of the conference
with all the participants, not only PC members.
Industrial panel
With powerful debaters from various industries
Dataset sharing service
Provides new datasets used by papers.
All datasets released as openly available
community resources. http://icwsm.cs.mcgill.ca
9. Resources
All the papers presented in the main
conference can be freely accessible from
http://www.aaai.org/Library/ICWSM/icwsm12contents.php
All the workshop papers are also free :
http://www.aaai.org/Library/Workshops/workshops-library.php
I gathered most tweets as to ICWSM 12,
freely accessible from
http://togetter.com/id/_akisato
10. Our presentation
Creating Stories : Social Curation of Twitter
Messages
Curated lists = supervised corpora for analyzing
microblog messages
http://www.brl.ntt.co.jp/people/akisato/socialweb1.html
11. 面白かった発表 1
The Livehoods Project: Utilizing Social Media
to Understand the Dynamics of a City
Won the Best Paper Award
Twitterタイムラインから取れる
位置情報(tweets with geotags, 4sq etc.)から,
かなり局所的な地域の特性の変化が掴める.
URL: http://livehoods.org
Twitter ID: @livehoods
20. 面白かった発表 その他羅列1
Crossing Media Streams with Sentiment:
Domain Adaptation in Blogs, Reviews and
Twitter
Sentiment analysisをTwitterだけでやるの
無理だから,reviewやblogを教師に使う.
Exploring Social-Historical Ties on Location-
Based Social Networks
Foursquareもの.トピックと位置,両方使う.
階層Pitman-Yor過程によるモデル化
21. 面白かった発表 その他羅列2
The Emergence of Conventions in Online
Social Networks
Won the Best Paper Award
Twitterにおける「文法」らしきものは,基本
的にボトムアップにできあがってきたもの.
それを網羅的に検証.