Detecting illegal shipments in the global timber trade poses a massive challenge to enforcement agencies. The massive volume and complexity of timber shipments and obfuscations within international trade data, intentional or not, necessitates an automated system to aid in detecting specific shipments that potentially contain illegally harvested wood. To address these requirements we build a novel human-in-the-loop visual analytics system called TIMBERSLEUTH. TimberSleuth uses a novel scoring model reinforced through human feedback to improve upon the relevance of the results of the system while using an off-the-shelf anomaly detection model. Detailed evaluation is performed using real data with synthetic anomalies to test the machine intelligence that drives the system. We design interactive visualizations to enable analysis of pertinent details of anomalous trade records so that analysts can determine if a record is relevant and provide iterative feedback. This feedback is utilized by the machine learning model to improve the precision of the output.
Trade in wood and forest products spans the global supply chain. Illegal logging and associated trade in forest products present a persistent threat to vulnerable ecosystems and communities. Illegal timber trade has been linked to violations of tax and conservation laws, as well as broader transnational crimes. The United States is the largest importer globally of wood and forest products, such as pulp, paper, flooring, and furniture—importing $78 billion in 2021. Transaction-level data such as shipping container manifests and bills of lading provide a comprehensive data source that can be used to detect and disrupt trade that may be suspected of containing illegally harvested or traded forest products. Owing to the volume, velocity, and complexity of shipment data, an automated decision support system is required for the purposes of detecting suspicious forest product shipments. We present a proof of concept framework using machine learning and big data approaches—combining domain expertise with automation—to achieve this objective. We formulated the underlying machine learning problem as an anomaly detection problem and collected and collated forest sector-specific domain knowledge to filter and target shipments of interest. In this work, we provide the overview of our framework, with the details of domain knowledge extraction and machine learning models, and discuss initial results and analysis of flagged anomalous and potentially suspicious records to demonstrate the efficacy of this approach. The proof of concept work presented here provides the groundwork for an actionable and feasible approach to assisting enforcement agencies with the detection of suspicious shipments that may contain illegally harvested or traded wood.
3 April 2020 in Proceedings of the AAAI Conference on Artificial Intelligence
Developing algorithms that identify potentially illegal trade shipments is a non-trivial task, exacerbated by the size of shipment data as well as the unavailability of positive training data. In collaboration with conservation organizations, we develop a framework that incorporates machine learning and domain knowledge to tackle this challenge. Modeling the task as anomaly detection, we propose a simple and effective embedding-based anomaly detection approach for categorical data that provides better performance and scalability than the current state-of-art, along with a negative sampling approach that can efficiently train the proposed model. Additionally, we show how our model aids the interpretability of results which is crucial for the task. Domain knowledge, though sparse and scattered across multiple open data sources, is ingested with input of domain experts to create rules that highlight actionable results. The application framework demonstrates the applicability of our proposed approach on real world trade data. An interface combined with the framework presents a complete system that can ingest, detect and aid in the analysis of suspicious timber trades.
1 January 2019 in Socius: Sociological Research for a Dynamic World
The Fragile Families Challenge charged participants to predict six outcomes for 4,242 children and their families interviewed in the Fragile Families and Child Wellbeing Study. These outcome variables are grade point average, grit, material hardship, eviction, layoff and job training. The data set provided contained longitudinal survey and observational data collected on families and their children from birth to age 9. The authors used these data to create models to make predictions at age 15. The authors describe the imputation and modeling strategies that led them to make predictions ranked fifth and ninth in the material hardship and layoff categories, respectively. However, the results of the study are inconclusive with respect to increased predictive performance. The authors view this work as a first step toward organizing the Fragile Families missing data by exploiting the structure of the survey instruments.
As the field of multimedia computing has grown rapidly, so has the need for larger datasets[5] and increased modeling capacity. Navigating this complex landscape often necessitates the use of sophisticated tools and cloud architectures, which all need to be addressed before the actual research commences. Recently, AutoML, an innovation previously exclusive to tabular data, has expanded to encompass multimedia data. This development has the potential to greatly streamline the research process, allowing researchers to shift their focus from model construction to the core content of their problems. In doing so, AutoML not only optimizes resource utilization but also boosts the reproducibility of results. The aim of this tutorial is to acquaint the multimedia community with AutoML technologies, underscoring their advantages and their practical applications in the field.
Toxin classification of protein sequences is a challenging task with real world applications in healthcare and synthetic biology. Due to an ever expanding database of proteins and the inordinate cost of manual annotation, automated machine learning based approaches are crucial. Approaches need to overcome challenges of homology, multi-functionality, and structural diversity among proteins in this task. We propose a novel deep learning based method ProtTox, that aims to address some of the shortcomings of previous approaches in classifying proteins as toxins or not. Our method achieves a performance of 0.812 F1-score which is about 5% higher than the closest performing baseline.