We describe a system that automatically generates from a curated database a collection of short conventional publications—citation summaries—that describe the contents of various components of the database. The purpose of these summaries is to ensure that the contributors to the database receive appropriate credit through the currently used measures such as h-indexes. Moreover, these summaries also serve to give credit to publications and people that are cited by the database. In doing this, we need to deal with granularity—how many summaries should be generated to represent effectively the contributions to a database? We also need to deal with evolution—for how long can a given summary serve as an appropriate reference when the database is evolving? We describe a journal specifically tailored to contain these citation summaries. We also briefly discuss the limitations that the current mechanisms for recording citations place on both the process and value of data citation.

The ubiquitous use of machine learning algorithms brings new challenges to traditional database problems such as incremental view update. Much effort is being put in better understanding and debugging machine learning models, as well as in identifying and repairing errors in training datasets. Our focus is on how to assist these activities when they have to retrain the machine learning model after removing problematic training samples in cleaning or selecting different subsets of training data for interpretability. This paper presents an efficient provenance-based approach, PrIU, and its optimized version, PrIU-opt, for incrementally updating model parameters without sacrificing prediction accuracy. We prove the correctness and convergence of the incrementally updated model parameters, and validate it experimentally. Experimental results show that up to two orders of magnitude speed-ups can be achieved by PrIU-opt compared to simply retraining the model from scratch, yet obtaining highly similar models.

No abstract

No abstract

Forecasting on sparse multivariate time series (MTS) aims to model the predictors of future values of time series given their incomplete past, which is important for many emerging applications. However, most existing methods process MTS’s individually, and do not leverage the dynamic distributions underlying the MTS’s, leading to sub-optimal results when the sparsity is high. To address this challenge, we propose a novel generative model, which tracks the transition of latent clusters, instead of isolated feature representations, to achieve robust modeling. It is characterized by a newly designed dynamic Gaussian mixture distribution, which captures the dynamics of clustering structures, and is used for emitting time series. The generative model is parameterized by neural networks. A structured inference network is also designed for enabling inductive analysis. A gating mechanism is further introduced to dynamically tune the Gaussian mixture distributions. Extensive experimental results on a variety of real-life datasets demonstrate the effectiveness of our method.