Restoration of linear disturbances from oil-and-gas exploration in boreal landscapes: How can network models help?

In western Canada, decades of oil-and-gas exploration have fragmented boreal landscapes with a dense network of linear forest disturbances (seismic lines). These seismic lines are implicated in the decline in wildlife populations that are adapted to function in unfragmented forest landscapes. In particular, anthropogenic disturbances have led to a decline of woodland caribou populations due to increasing predator access to core caribou habitat. Restoration of seismic lines aims to reduce the landscape fragmentation and stop the decline of caribou populations. However, planning restoration in complex landscapes can be challenging because it must account for a multitude of diverse aspects. To assist with restoration planning, we present a spatial network optimization approach that selects restoration locations in a fragmented landscape while addressing key environmental and logistical constraints. We applied the model to develop restoration scenarios in the Redrock-Prairie Creek caribou range in northwestern Alberta, Canada, which includes a combination of caribou habitat and active oil-and-gas and timber extraction areas. Our study applies network optimization at two distinct scales to address both the broad-scale restoration policy planning and project-level constraints at the level of individual forest sites. We first delineated a contiguous set of coarse-scale regions where restoration is most cost-effective and used this solution to solve a fine-scale network optimization model that addresses environmental and logistical planning constraints at the level of forest patches. Our two-tiered approach helps address the challenges of fine-scale spatial optimization of restoration activities. An additional coarse-scale optimization step finds a feasible starting solution for the fine-scale restoration problem, which serves to reduce the time to find an optimal solution. The added coarse-scale spatial constraints also make the fine-scale restoration solution align with the coarse-scale landscape features, which helps address the broad-scale restoration policies. The approach is generalizable and applicable to assist restoration planning in other regions fragmented by oil-and-gas activities.

Utilizing deep learning and graph mining to identify drug use on Twitter data.

BACKGROUND: The collection and examination of social media has become a useful mechanism for studying the mental activity and behavior tendencies of users. Through the analysis of a collected set of Twitter data, a model will be developed for predicting positively referenced, drug-related tweets. From this, trends and correlations can be determined. METHODS: Social media data (tweets and attributes) were collected and processed using topic pertaining keywords, such as drug slang and use-conditions (methods of drug consumption). Potential candidates were preprocessed resulting in a dataset of 3,696,150 rows. The predictive classification power of multiple methods was compared including SVM, XGBoost, BERT and CNN-based classifiers. For the latter, a deep learning approach was implemented to screen and analyze the semantic meaning of the tweets. RESULTS: To test the predictive capability of the model, SVM and XGBoost were first employed. The results calculated from the models respectively displayed an accuracy of 59.33% and 54.90%, with AUC's of 0.87 and 0.71. The values show a low predictive capability with little discrimination. Conversely, the CNN-based classifiers presented a significant improvement, between the two models tested. The first was trained with 2661 manually labeled samples, while the other included synthetically generated tweets culminating in 12,142 samples. The accuracy scores were 76.35% and 82.31%, with an AUC of 0.90 and 0.91. Using association rule mining in conjunction with the CNN-based classifier showed a high likelihood for keywords such as "smoke", "cocaine", and "marijuana" triggering a drug-positive classification. CONCLUSION: Predictive analysis with a CNN is promising, whereas attribute-based models presented little predictive capability and were not suitable for analyzing text of data. This research found that the commonly mentioned drugs had a level of correspondence with frequently used illicit substances, proving the practical usefulness of this system. Lastly, the synthetically generated set provided increased accuracy scores and improves the predictive capability.