精力有限,希望同学们补充“摘要”,以及划重点(加粗)
- Document and topic models
- One Theme in All Views: Modeling Consensus Topics in Multiple Contexts
- Representing Documents Through Their Readers
- 以往文章研究的是给定文章model用户,这篇文章要给定用户model文章,比如一个用户的profile里有‘素食’,那么可能这个人分享的文章就和‘素食’有关。
- 构建了一个单词和类别的dictionary,然后利用单词线性组合成句子,来model文章
- 以往文章研究的是给定文章model用户,这篇文章要给定用户model文章,比如一个用户的profile里有‘素食’,那么可能这个人分享的文章就和‘素食’有关。
- Text-Based Measures of Document Diversity
- Diversity Maximization Under Matroid Constraints
- 文章聚类显示时,每个分类要多样性展示
- 定义距离函数,要求pair-wise的距离大
- 每个子分类只能有一个选入
- 文章聚类显示时,每个分类要多样性展示
- One Theme in All Views: Modeling Consensus Topics in Multiple Contexts
- Social media
- Connecting Users across Social Media Sites: A Behavioral-Modeling Approach
- cross-media user identication problem
- cross-media user identication problem
- Automatic selection of social media responses to news
- 在所有refer某一篇article的‘tweet’中寻找最有代表性的,可以是视频之类的
- 在所有refer某一篇article的‘tweet’中寻找最有代表性的,可以是视频之类的
- Estimating Unbiased Sharer Reputation via Social Data Calibration
- not influential user, but 优质用户
- not influential user, but 优质用户
- Linking Named Entities in Tweets with Knowledge Base via User Interest Modeling
- tweet entity linking
- graph model, each user has a underline topic distribution
- tweet entity linking
- Connecting Users across Social Media Sites: A Behavioral-Modeling Approach
- Big data frameworks
- 暂不关注
- Graph mining
- Selective Sampling on Graphs for Classification
- online learning
- graph sampling
- online learning
- Selective Sampling on Graphs for Classification
- Classification
- Density-Based Logistic Regression
- nonlinear logistic regression
- map the data to a feature space based on kernel density estimation
- nonlinear logistic regression
- MILS: Multi-Instance Learning from Multiple Information Sources
- Querying Discriminative and Representative Samples for Batch Mode Active Learning
- SVM_{pAUC}^{tight}: A New Support Vector Method for Optimizing Partial AUC Based on a Tight Convex Upper Bound
- Density-Based Logistic Regression
- Healthcare and bioinformatics
- 暂不关心
- Recommender systems
- Learning to question: Leveraging user preferences for shopping advice
- 生成一棵决策树,但是每个node是一个非专业性的问题(如买电脑,问是否是学生?),在每个node都会生成一个ranklist,可以在任意层退出
- 生成一棵决策树,但是每个node是一个非专业性的问题(如买电脑,问是否是学生?),在每个node都会生成一个ranklist,可以在任意层退出
- Active Learning and Search on Low-Rank Matrices
- LCARS: A Location-Content-Aware Recommender System
- Learning to question: Leveraging user preferences for shopping advice
- Temporal/social influence
- Discovering Latent Influence in Online Social Activities via Shared Cascade Poisson Processes
- shared cascade Poisson processes for discovering latent influence from multiple sequences of item adoption events
- shared cascade Poisson processes for discovering latent influence from multiple sequences of item adoption events
- STRIP: Stream Learning of Influence Probabilities
- Fast Structure Learning in Generalized Stochastic Processes with Latent Factors
- Discovering Latent Influence in Online Social Activities via Shared Cascade Poisson Processes
- Sparse learning
- Robust Sparse Estimation of Multiresponse Regression and Inverse Covariance Matrix via the L2 distance
- Exact Sparse Recovery with L0 Projections
- Robust Principal Component Analysis via Capped Norms
- Robust Sparse Estimation of Multiresponse Regression and Inverse Covariance Matrix via the L2 distance