Skip to content

GyaneshSamanta/Jina.Ai-blog

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 

Repository files navigation

About Jina.Ai

An easier way to build neural search in the cloud

But wait, what does it even mean?

Context

Here we are gonna talk about search engines and the way search engines "search" for stuff when we type in something.

For people who are unfamiliar with the term search engine, it is a software system that is designed to carry out web searches. They search the world wide web in a way to find the particular keywords you searched for.

In the early days, due to the limitations of data available to us (and the processing speeds of the silicon), search engines would look for the exact keywords you typed in.

For example, if you typed in America as Amrica it'd most likely send out a zero search result.

So far, we've listed three particular problems-

  • Limited data available to the public
  • Speed of the silicon chips being very limited
  • the searching algorithms used by search engines being pretty basic

As the amount of data available increased and the speed of the silicon chips caught up to process millions of instructions per second, we needed a better way to search the internet.

Neural Networks

Allow me to take you back on a journey to 1959 when Arthur Samuel, an American IBMer coined the term machine learning. Arthur Samuel was a pioneer in the field of computer gaming and artificial intelligence. But here we're gonna focus on the background knowledge for ML that we need to understand before coming to neural networks.

Machine Learning

Here are 4 key things that we'll be comparing in order to understand AI, ML, Neural Networks and Deep Learning.

comparision

As you can see in this diagram, in simple terms we can consider Neural Networks to be a part of Machine Learning.

So what exactly is neural networks?

Defining Neural Networks

Neural networks mimic the human brain through a set of algorithms, At a basic level, it is comprised of inputs, weights, a bias or threshold and an output.

The algebraic formula would look like this -

Formula

We train the neural networks to identify and predict the pattern by guessing.

For example cofee can be predicted to be coffee. This provides more flexibility compare to the basic rule based system.

Jina.ai

An easer way to build in the cloud.

Jina provides the design pattern for building neural search on the cloud.

some key features of Jina-

  • Universal Search
  • Time Saver
  • Full Stack ownership
  • Fast and cloud ready
  • first class AI models

To learn more about getting started with Jina and use Jina for yourself use the link below

GitHub Repository

To test how Jina can be used to create and implement a COVID-19 Chatbot use this link

Build Covid chatbot with Jina

Releases

No releases published

Packages

No packages published