Crime Event Embedding with Unsupervised Feature Selection

  title={Crime Event Embedding with Unsupervised Feature Selection},
  author={Shixiang Zhu and Yao Xie},
  journal={ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  • Shixiang Zhu, Yao Xie
  • Published 2019
  • Computer Science, Mathematics
  • ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
We present a novel event embedding algorithm for crime data that can jointly capture time, location, and the complex free-text component of each event. The embedding is achieved by regularized Restricted Boltzmann Machines (RBMs), and we introduce a new way to regularize by imposing a `1 penalty on the conditional distributions of the observed variables of RBMs. This choice of regularization performs feature selection and it also leads to efficient computation since the gradient can be computed… Expand
Crime Linkage Detection by Spatio-Temporal-Textual Point Processes
This work proposes a new modeling and learning framework for detecting linkage between crime events using spatio-temporal-textual data, which are highly prevalent in the form of police reports, and captures the notion of modus operandi, by introducing a multivariate marked point process and handling the complex text jointly with the time and location. Expand
Spatial-Temporal-Textual Point Processes with Applications in Crime Linkage Detection
Crimes emerge out of complex interactions of human behaviors and situations. Linkages between crime events are highly complex. Detecting crime linkage given a set of events is a highly challengingExpand
Assessing GAN-based approaches for generative modeling of crime text reports
This paper evaluates the ability of generative adversarial network models to represent crime text data and generate realistic crime reports and compares four state of the art GAN algorithms in terms of quantitative metrics such as coherence, embedding similarity, negative log-likelihood, and qualitatively based on inspection of generated text. Expand
Spatial-Temporal-Textual Point Processes for Crime Linkage Detection
  • Shixiang Zhu, Yao Xie
  • Mathematics, Computer Science
  • 2019
The joint modeling of space, time, and text information enhances crime linkage detection compared with the state-ofthe-art, and the learned spatial dependence from data can be useful for police operations. Expand
Reinforcement Learning of Spatio-Temporal Point Processes
A novel and flexible semi-parametric spatial-temporal point processes model is introduced, by combining spatial statistical models based on heterogeneous Gaussian mixture diffusion kernels, whose parameters are represented using neural networks. Expand
An approach for understanding offender modus operandi to detect serial robbery crimes
An approach for detecting serial robbery crimes based on understanding offender M.O. by integrating crime process information, which shows that the addition of crime process Information can substantially improve the effect of detecting serial crimes. Expand
Latent Embeddings of Point Process Excitations
This work uses the Hawkes framework as a vessel for embedding arbitrary event types in a new latent space and demonstrates that learning the embedding alongside a point process model uncovers the coherent, rather than spurious, interactions. Expand
Spatio-Temporal Point Processes with Attention for Traffic Congestion Event Modeling
This work presents a novel framework for modeling traffic congestion events over road networks based on new mutually exciting spatio-temporal point process models with attention mechanisms and neural network embeddings, and introduces a novel attention-based mechanism based on neural networks embedding for the point process model. Expand
Imitation Learning of Neural Spatio-Temporal Point Processes
A novel Neural Embedding Spatio-Temporal point process model for spatio-temporal discrete event data is presented and an efficient imitation learning (a type of reinforcement learning) based approach for model fitting is developed. Expand


Crime Incidents Embedding Using Restricted Boltzmann Machines
  • Shixiang Zhu, Yao Xie
  • Computer Science, Mathematics
  • 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
  • 2018
Experiments over several series of related crime incidents hand labeled by the Atlanta Police Department reveal the promise of the Gaussian-Bernoulli Restricted Boltzmann Machine embedding methods. Expand
Finding Patterns with a Rotten Core: Data Mining for Crime Series with Cores
The goal is to assist crime analysts by providing automated tools for discovering crime series from within a database of crimes by proposing a subspace clustering method that relies on a key hypothesis that each crime series possesses at least one core of crimes that are very similar to each other, which can be used to characterize the modus operandi of the criminal. Expand
TrioVecEvent: Embedding-Based Online Local Event Detection in Geo-Tagged Tweet Streams
Crowdourcing is used to evaluate TrioVecEvent, a method that leverages multimodal embeddings to achieve accurate online local event detection and introduces discriminative features that can well characterize local events. Expand
Event2vec: Learning Representations of Events on Temporal Sequences
An algorithm called event2vec is proposed that can be used to improve the quality of health-care without any additional burden and is totally unsupervised without the help of expert knowledge. Expand
Visualizing Data using t-SNE
We present a new technique called “t-SNE” that visualizes high-dimensional data by giving each datapoint a location in a two or three-dimensional map. The technique is a variation of StochasticExpand
Recurrent Marked Temporal Point Processes: Embedding Event History to Vector
The Recurrent Marked Temporal Point Process is proposed to simultaneously model the event timings and the markers, and uses a recurrent neural network to automatically learn a representation of influences from the event history, and an efficient stochastic gradient algorithm is developed for learning the model parameters. Expand
Sparse Feature Learning for Deep Belief Networks
This work proposes a simple criterion to compare and select different unsupervised machines based on the trade-off between the reconstruction error and the information content of the representation, and describes a novel and efficient algorithm to learn sparse representations. Expand
Context Selection for Embedding Models
This paper shows that conditioning on all the elements in the context is not optimal, and model the probability of the target conditioned on a learned subset of the elementsIn the context, and uses amortized variational inference to automatically choose this subset. Expand
Exponential Family Harmoniums with an Application to Information Retrieval
An alternative two-layer model based on exponential family distributions and the semantics of undirected models is proposed, which performs well on document retrieval tasks and provides an elegant solution to searching with keywords. Expand
Distributed Representations of Words and Phrases and their Compositionality
This paper presents a simple method for finding phrases in text, and shows that learning good vector representations for millions of phrases is possible and describes a simple alternative to the hierarchical softmax called negative sampling. Expand