Milvus — Vector DB, Embeddings and Practical Significance

What is Unstructured data

Unstructured data refers to information that isn’t organized in a predefined manner — this includes text, audio, video, social media content, and so on. For instance, a webpage full of textual content, a library of songs, or a collection of videos — all of these are examples of unstructured data. Managing and making sense of such data can be a complex task, given its diverse nature and absence of a clear structure.

This is where vector databases come into play. They offer a powerful solution to handle unstructured data by converting this data into a numerical form, a process often referred to as vectorization.

Pic from Kensho Communications

Simple Vector DB Example

To illustrate, let’s consider a simple example of text data. Suppose we have a set of customer reviews for a product in textual form. It’s a form of unstructured data as it doesn’t adhere to a specific format or structure. To analyze this data using machine learning algorithms, we need to convert these reviews into a numerical form.