How to architect/design a knowledge base to solve issues from its history analysis?

I have a ticketing system (lets say JIRA or similar) for my application to file an issue of my application. Now my requirement is to build a knowledge base in a way so that I can predict the solution of any similar issues in future by churning that knowledge base.

To explain further, the knowledge base would give me how many times this kind of issues have arisen in past and what have been the root cause of it in most of the time (lets say 80% time). This way the repository should have an analysis of each and every issue and its possible root cause plus many other relevant information about the issue.

Just to start off to build such a knowledge base, I need to know following things:

1. What is the most commonly used technology/mechanism available to achieve this ?
2. How do I need to architect/design a system to be able to serve this kind of requirement?
3. Does it require to learn any particular language/database ?

I request community experts to enlighten me with the required information and pointers to give me a starting point at least in this direction.

Thanks.

I would suggest against a ‘reinvent the wheel’ approach. There are perfectly good tools out there that achieve your required use cases. Look at ServiceNow or Desk.com as CRM for tickets, or if you just want a Wiki that integrates with Jira, look at Atlassian’s wiki. You can also generate reports from Jira itself, not sure why anyone would want to build his own when there are such great tools out there.