Skip to content

Roadmap

Kristin Antin edited this page Apr 18, 2019 · 17 revisions

Improve relationships

Manage display of relationships

  • Be able to regroup scattered connections in a hub.
  • Connections can also be rendered in the side panel, both in library and in connections view (that would be, the connections of something connected).
  • Admins can define a default view (grouping, filtering and sorting)
  • Be able to group by entity type and connection type, instead of hubs.

Performance

  • Improve performance for high volume of relationships.

Relationship metadata

  • Admin interface for defining relationship metadata fields.
  • Filtering by connection metadata.

Create entities and docs from relationships view as a better user workflow (2-4d):

  • The search panel allows for new entity creation
  • Entities can be edited and deleted from relationship view
  • Allow publishing from side panel
  • Be smart when editing/deleting so the user experience is smooth
  • Do not duplicate information in the stores

Inherit metadata from relationship (1w)

  • Be able to select a target field in template configuration
  • Index the metadata of the selected field

Text connections revamp

  • Text references from pages, side panel in pages + several improvements

Be able to create new relationship types from the relationships tree.

Library and document search improvements

Form for data intake (2-4d)

  • End-point. Has an array contract
  • Can create multiples entities
  • Make metadata form renderable in pages

Advanced search

  • More accessible documents.

    • Be able to render a text version of the document. Needed for screen readers, search engines and to have proper copy/paste.
    • Detach our text references from a particular PDF.js version. So we can keep that library up to date.
    • Have a pre-render so documents load faster.
    • Have a document thumbnail with the first page.
  • Better way to deal with default properties (title/date added)

  • OCR integration

  • Conversion for other formats

Arbitrary order for select/multiselect (thesauri) items: (2d)

  • Users can sort thesaurus in the thesaurus page

  • We will respect the order that the users configured

  • A "sort alphabetically" flag will be passed from thesauris to the fields

  • This order should be rendered in cards, side panels, and entity view.

  • Support for video and image document types or field types

  • Unpublished documents workflow

Data Visualisation: Graphs, charts, maps

Geolocation (1w)

  • Metadata field relationship can handle geolocation points
  • Name to location resolution (2-3d) - Lookup to a remote service

Graphs

  • Build a user interface to allow admins to create their own graphs and charts.
  • Support dynamic lengths based on dates as fields and data viz (1-2w)

Table view (5-7d)

  • Show all properties/only show in cards toggle
  • Use a tabular component (research)
  • Add a button to the UI

Timeline dataviz (1-2w)

  • Page component
  • Library visualization (vertical TL)
  • Improve styles
  • Some configuration (for the pages component)

Improved security features

Audit log (4-6d):

  • Log edit/delete calls
  • Present data in a user understandable way
  • Log the post data only in database (users can't see it)
  • Track user and dates
  • Action type, model
  • Pagination

Security audit

Permissions, roles, level of access

  • Document ownership, different levels of access/sharing (light permissions)

Machine learning

  • Automations: machine learning and other classifiers

Long term/ideas

  • Move document and entity types to "Templates".

  • Have a type of template selector (document, video, entity, image, audio...).

  • Matadata from document text workflow + template creation from anywhere

  • Federation/syncing

  • Advanced search

  • Hybrid persistence graph/doc

  • Comments/annotation - not sure about this one

  • Crawl and import existing documents on a website

Data import/export

CSV export (2-3d) https://github.com/huridocs/uwazi/issues/958

  • Export only values (no cross-referencing)
  • All records
  • Depends on table view

CSV import entities (1-2w) https://github.com/huridocs/uwazi/issues/2310 (this is being developed now)

  • Test-run to check for errors
  • Thesauris and relationship entities are imported apart
  • 1 CSV for 1 entity type
  • Automatically transform data into the target field
  • Template needs to be pre-configured
  • 1 single CSV with columns per language for thesauris
  • Name lookup to avoid duplicates on creation (2-3d)
Clone this wiki locally