The Most Critical AI Tools for Investigating Companies and People
January 18, 2024 | by Christina McEvoy
This is part 2 of our 3-part mini-blog series.
To enhance their effectiveness in performing ongoing due diligence for their clients, investigators must maintain proficiency in the use of an array of AI tools. At Hetherington Group, we use a suite of AI tools to monitor social media, editorial media, data breaches, and the dark web. Monitoring tools can be as simple as Google Alerts and RSS dashboards to complex facial recognition software. While the variety of AI tools is extensive, the common tools most investigators should master include the following:
- Web Scraping Tools – tools such as Beautiful Soup pull data from HTML and XML files
- Text Analysis and Natural Language Processing (NLP) Tools – these tools assist in mining human language data such as tagging, parsing and tokenization
- Image and Video Analysis Tools – tools such as Google Cloud Vision API and Microsoft Azure Computer Vision API enable image labeling, face detection and optical character recognition (OCR)
- Social Media Analysis AI Tools – tools such as Tweepy and facial recognition libraries such as OpenCV or dlib enable recognition of faces in images and videos
- Data Visualization Tools – An array of tools such as Seaborn and Tableau Public allow for creation of static, animated/interactive and shareable data visualizations
- Machine Learning Tools – numerous machine learning libraries enable investigators to navigate algorithms for classification, regression, clustering and dimensionality reduction
- Link Analysis Tools – these tools enable visualization of relationships and links between entities in a graphical format
- Automation and Scripting Tools – these resources include Python scripting to automate data retrieval from websites
- Search Engines and Aggregators – the Google Custom Search Engine (CSE) is one example that allows investigators to create custom search engines for highly targeted information retrieval
While AI tools can help speed up efficiency, expand the realm of what is possible in an investigation, and allow you to take on more complex investigations, it can never fully replace human investigators. Part 3 of this blog series will discuss why.