Machine Learning
Through the use of Machine Learning I have developed tools to: predict house prices, intelligently price Swimply offerings, predict AirBnB prices, classify email senders by writing style, find material misrepresentations, and analyze social media.
- Machine Learning Algorithm to Predict Peaks and Valleys of Weekly SPY Options
- Reddit News Sentiment Analysis
- House Price Prediction
Python Tools
Since 2017 I have built many Python tools to: archive websites, alert end users, crawl javascript heavy websites, analyze advertisements, scrape social media, collect open source intelligence, and much more. Select tools include:
- Selenium Triple Archiver (Also available as an API running via a dockerized AWS Lambda service.)
- Selenium Tool to Classify Dog Breed By Image via Microsoft Visual Search
- Google Ads Results Scraper (Built to archive Google Ads for a future machine learning project to identify suspicious ads and websites.)
- Selenium Facebook Page Link Scraper (Used to find Facebook page links by topic for competitor analysis, suspicious activity monitoring, sentiment analysis, industry growth, etc.)
- Additional tools automate Facebook actions and collect and archive Facebook posts, images, text, messages, comments, etc. The primary purpose of these tools is to collect open source intelligence to be used for anomaly detection, competitor analysis, sentiment analysis, or machine learning.
- Javascript Friendly Web Crawler (Currently used to download PDF and images from Philippines government websites because the Philippines government posts many important documents in image or scanned-PDF format. These documents are not text searchable. However, by downloading these documents via a javascript friendly scraper it is easy to search and analyze the text in these documents via Open Semantic Search.)
Data Visualizations
Telling a story through visualizations and interactive programs can be more important than the data itself.
- Know Your Agency Visualizer, as seen on South China Morning Post.
- Money Lender, Third Party Money Lender, Agency, MSO, and TCSP Network Map
- 2021 POEA Case Data Visualization
Presentations
Throughout my career I have conducted many presentations and webinars, linked below are two of my most recent.
Legal and Policy Contributions
While working at Migrasia, I built numerous tools to automate tasks, collect & analyze open source intelligence, and identify instances of activity that were suspected to contravene the laws of Hong Kong SAR. Certain elements of my work are summarized in an overview presentation.
- Leveraged open source intelligence and anomaly detection to identify a ponzi-like agency scheme. After detection I worked with authorities to collect evidence, arrange witness interviews, and present the case for prosecution.
- Organized one of the groups that lobbied the Hong Kong SAR government to increase the statute of limitations and penalties for overcharging offenses by employment agencies.
Academic Work
While studying BAITM at WGU I completed a number of projects and received multiple awards, including: