Features
Real-time hot topic detection with sliding time window
WOLDA algorithm for streaming topic modeling
Sentiment analysis combining SVM and logistic regression
Lexicon-based sentiment features from 5 sentiment lexicons
Representative tweet extraction via KL-mean, cosine distance, maximum entropy
Dynamic vocabulary management (handles new words, slang)
Integration with Twitter Streaming API (v1.1)
Topic evolution treemap visualization
Bubble chart visualization of topic-word distributions
Sunburst chart for hierarchical topic-word display
Hashtag frequency bar and trend charts
Geolocation heatmap visualization
Sliding window forgetting to detect new topics
Resource-efficient incremental model updates
POS tagging with CMU ArkTweetNLP
Multi-sensor core scanning: RGB, XRF, hyperspectral, LiDAR
LIBS-based detection of REEs and light elements (via Lumo Analytics)
AI-powered core logging on Digital Core Table
Resource modeling with RMSP integration
Drill Hole Optimizer for mine planning
Sub-48-hour turnaround time
4x faster than manual core logging
Over 400% project acceleration
End-to-end workflow from scanning to modeling
Domain expertise throughout mining cycle
High-fidelity data capture and analytics
Consistent logging with fewer errors
Professional consulting and training services
Decision engineering for critical mineral exploration
Cloud-based digital core collaboration
