Opened 8 years ago
Closed 7 years ago
#855 closed enhancement (wontfix)
Make available selected facets to LR switchboard
| Reported by: | Claus Zinn | Owned by: | Twan Goosen |
|---|---|---|---|
| Priority: | major | Milestone: | VLO-4.2 or later |
| Component: | VLO web app | Version: | |
| Keywords: | Faceted Browser - User Intent, Switchboard | Cc: | Dieter Van Uytvanck, teckart@informatik.uni-leipzig.de |
Description
CLARIN-PLUS will integrate the LR switchboard to the VLO. The switchboard suggests, for any given resource a user has found with the help of the faceted browser, tools that can process the resource. Example: given the resource is in German, and has a mimetype equal to plain/text, the switchboard could propose a German POS tagger or dependency parser. In case such metadata about the resource cannot be easily read from the metadata (by the switchboard), it'd be nice to have the faceted browser making available its current state to the switchboard: which facets have been selected, capturing thus the user's intention about the resource (s)he's after.
In short: make available all facets a user has selected to the LR switchboard. Suggested format: a JSON structure, listing all facets selected with their value, e.g., { language: "German", format: "text/xml" }.
Change History (10)
comment:1 Changed 8 years ago by
| Cc: | teckart@informatik.uni-leipzig.de added |
|---|
comment:2 Changed 8 years ago by
| Type: | defect → enhancement |
|---|
comment:3 Changed 8 years ago by
| Milestone: | → VLO-4.0 or later |
|---|
I'm putting this on the 'foreseeable future' milestone, as I'm not sure when this integration is expected to happen but I assume it's going to be sometime in the next 3-12 months - correct?
To be assigned to a more concrete milestone in the short term as far as I'm concerned.
comment:4 follow-up: 5 Changed 8 years ago by
Yes, it has to happen during the browser session. In the future VLO, a switchboard icon will be associated with each resource of interest (the one users have found doing full-text and faceted search). When the user clicks on the switchboard icon, the switchboard is invoked with the following information: the resource in question, and the VLO state (facets selected).
Timeline: by the end of Oct 2016, the switchboard will be connected to the VLO and FCS (final, stable). There are timelines for beta-testing and release testing earlier in the summer. So it is all quite tight. When is the VLO4.0 release scheduled for?
comment:5 Changed 8 years ago by
Replying to claus.zinn@…:
Yes, it has to happen during the browser session. In the future VLO, a switchboard icon will be associated with each resource of interest (the one users have found doing full-text and faceted search). When the user clicks on the switchboard icon, the switchboard is invoked with the following information: the resource in question, and the VLO state (facets selected).
Ok, so the icon should trigger some request to the switchboard that includes the VLO query state. That is fairly trivial as far as the VLO is concerned.
Timeline: by the end of Oct 2016, the switchboard will be connected to the VLO and FCS (final, stable). There are timelines for beta-testing and release testing earlier in the summer. So it is all quite tight. When is the VLO4.0 release scheduled for?
Version 4.0 is scheduled for release in May. So this is probably going to be 4.1 then (we don't have an exact schedule that far ahead yet, but summer-early autumn sounds likely).
comment:7 Changed 8 years ago by
@claus: is this ticket still in sync with the requirements and interface/implementation of the LRS?
comment:9 follow-up: 10 Changed 8 years ago by
Currently, the VLO invokes the LRS via a URL that encodes the resource's URL, its mimetype and its language, and currently this is the only information that the LRS uses to find applicable tools. In the future, the LRS "may" take into account different information, but we could encode this information then also in the URL. I suspect that the user's intent (as captured by the facet selection) corresponds one-to-one with the information that is URL-encoded, so for the time being, we may not need this feature, but keep it in mind for the future.
comment:10 Changed 7 years ago by
| Resolution: | → wontfix |
|---|---|
| Status: | new → closed |
Replying to claus.zinn@…:
Currently, the VLO invokes the LRS via a URL that encodes the resource's URL, its mimetype and its language, and currently this is the only information that the LRS uses to find applicable tools. In the future, the LRS "may" take into account different information, but we could encode this information then also in the URL. I suspect that the user's intent (as captured by the facet selection) corresponds one-to-one with the information that is URL-encoded, so for the time being, we may not need this feature, but keep it in mind for the future.
When the need arises, we will create a new ticket. Closing this one to prevent pollution of the ticket lists.
This would be fairly easy to do. Would this data be posted to the switchboard endpoint or do you have something else in mind, e.g. would the switchboard like to pull this data from the VLO somehow? I assume the former, which would also be the easiest to achieve as in the VLO no state is stored outside the user's session.