International Journal on Artificial Intelligence Tools

Social networks, e.g. Twitter, have been proved to be almost real-time systems for spreading information, that provide a valuable information channel in emergencies, e.g. disasters. This paper presents a framework designed to distill actionable tweets. The framework tackles the diversity, large volume, and noise of tweets for providing users live information for quick responses. To do that, our framework first retrieves a large number of tweets to ensure the diversity. It next removes irrelevant and indirect tweets for reducing the volume, divides informative tweets into predefined classes for quick navigation, and groups tweets in a class into topics to preserve the diversity. Finally, it ranks tweets in each topic to extract important tweets for the user’s quick scan. For ranking, the framework utilizes event extraction to enrich the semantics and reduce the noise of tweets. After that, the framework builds event graphs for ranking to find out important tweets. To validate the efficiency of our framework, we took Twitter as a case study. Experimental results on five disaster datasets show that our framework achieves promising results compared to strong methods in disaster scenarios.

Topological data analysis is a relatively new branch of machine learning that excels in studying high-dimensional data, and is theoretically known to be robust against noise. Meanwhile, data objects with mixed numeric and categorical attributes are ubiquitous in real-world applications. However, topological methods are usually applied to point cloud data, and to the best of our knowledge there is no available framework for the classification of mixed data using topological methods. In this paper, we propose a novel topological machine learning method for mixed data classification. In the proposed method, we use theory from topological data analysis such as persistent homology, persistence diagrams and Wasserstein distance to study mixed data. The performance of the proposed method is demonstrated by experiments on a real-world heart disease dataset. Experimental results show that our topological method outperforms several state-of-the-art algorithms in the prediction of heart disease.

Time series change point detection can identify the locations of abrupt points in many dynamic processes. It can help us to find anomaly data in an early stage. At the same time, detecting change points for long, periodic, and multiple input series data has received a lot of attention recently, and is universally applicable in many fields including power, environment, finance, and medicine. However, the performance of classical methods typically scales poorly for such time series.

In this paper, we propose CPMAN, a novel prediction-based change point detection approach via multi-stage attention networks. Our model consists of two key modules. First, in the time series prediction module, we employ the multi-stage attention-based networks and integrate the multi-series fusion mechanism. This module can adaptively extract features from the relevant input series and capture the long-term temporal dependencies. Secondly, in the change point detection module, we use the wavelet analysis-based algorithm to detect change points efficiently and identify the change points and outliers. Extensive experiments are conducted on various real-world datasets and synthetic datasets, proving the superiority and effectiveness of CPMAN. Our approach outperforms the state-of-the-art methods by up to 12.1% on the F1 Score.

Humans are capable of understanding the knowledge that is included in technical documents automatically by consciously combining the information that is presented in the document’s individual modalities. These modalities are mathematical formulas, charts, tables, diagram images and etc. In this paper, we significantly enhance a previously presented technical document understanding methodology³ that emulates the way that humans also perceive information. More specifically, we make the original diagram understanding methodology adaptive to larger architectures with more complex structures and modules. The overall understanding methodology results in the generation of a Stochastic Petri-net (SPN) graph that describes the system’s functionality. Finally, we conclude with the introduction of the hierarchical association of different diagram images from the same technical document. This processing step aims to provide a holistic understanding of all illustrated diagram information.

One of the most sought-after research areas in object detection is pedestrian detection owing to its applications especially in automated surveillance and robotics. Traditional methods use hand-crafted features to characterize pedestrians. In this work, we have pro-posed a new hand-crafted feature extraction method that concatenates shape, color and texture features; which is then classified by using Support Vector Machine (SVM). As in recent years, deep learning models such as Convolutional Neural Networks (CNNs) have become an eminent state of the art in detection challenges, which unlike the manually designed feature extraction mechanism, results in more accuracy. Therefore, we have also proposed a CNN network, a modification of the pre-trained ResNet-18 named as Multi-layer Feature Fused-ResNet (MF2-ResNet). We have used the proposed modification for (1) feature extraction; which is then classified by using Support Vector Machine (SVM); (2) End-to-End feature extraction and classification by the CNN network and (3) MF2-ResNet based Faster-RCNN to include region proposals in the detection pipeline. To evaluate the proposed method, it is compared with existing pre-trained CNNs. The MF2-ResNet based Faster R-CNN is compared with state-of-the-art detection methods. Three benchmark pedestrian datasets are used in this work: INRIA, NICTA and Daimler.