Currently, the report is a static file - shown as generated.

It would be useful to allow users to sort, filter and group the data as they wish.

Grouping would also sum the story points / effort for given group and filters.

These properties are a good candidates for grouping and filtering:

• Originating location (file)
• Originating app module/submodule/...
• Originating java package
• Which rule found the item
• Technologies involved

Appearantly:
1) Some of the properties are related as a tree (e.g. app module/submodule/ ...; package hierarhy. This could need tree-ish structures in the data.
2) Properties are specific to rulesets. Which means, a generic way to describe properties groupable/filterable is needed.

Possible solutions are:
1) Data as HTML tables, filtered by CSS (display:...), grouped in advance.
2) Data in JSON, filtered and grouped by JavaScript (libs).
3) Data in GraphSON, processed by JS library like HeliosJS which has Gremlin-like query fluent API. http://entrendipity.github.io/helios.js/
4) Data in XML, processed by XSLT (browser support).
5) Data in SQLite or similar, processed by browser support for local SQL database.
6) Data left in Titan, Rexter server started, data queried by Rexter REST API.

For the beginning, 1) could give some progress at low cost.
2) is portable but slow.
3) is porable but slow.
4) is portable and could be relatively fast.
5) is not that much portable, only recent browsers support local dbs.
6) was rejected, at least for now.