Vector search and NLP

Vector searching to power semantic or similarity searching is another approach that is starting to gain in popularity. In the simplest terms, it accomplishes this by converting the data to vectors and then leveraging one of several algorithms to determine nearest neighbors.

Fig. 89 Overview of vector searching

Searching for maliciousness with semantic context potentially opens up detection approaches that were impossible with just text based searching.

Natural Language Processing (NLP) is a specific approach to understanding human language, and can work in various ways, but common approaches include rule-based or statistical ML algorithms. Leveraging vector search on top of NLP can allow for providing deeper insights into the data.