Is there a way to log user actions at the dashboard page/tab level vs the dashboard on the whole, in Spotfire

[Vignesh Shekhar]

I'm looking to fetch the number of user views and the time spent on each dashboard page/tab within a dashboard.(my dashboard has about 8 pages/tabs). Is there a way to achieve this Note: Action logging is turned on at the server level and only logs user actions at the dashboard level overall and not granular at the page/tab specifc level.

Thanks in advance!

[Shandilya Peddi]

You can use the "set_page" action to look into when the user visited a sepecific page in the analysis

https://docs.tibco.com/pub/spotfire_server/10.9.0/doc/html/TIB_sfire_ser...

[Amit Nex]

@Shandilya Peddi - Does this also work in case the navigations are through action controls?

[David Boot-Olazabal]

Hi Amit,

Yes, this also works when using a action control button to navigate to another page.

Assuming that you have activated the actionlog, you could easily check the outcome of navigating to pages through action controls in the actionlog table (directly on the database or via the user_actions dxp file in Spotfire.

Kind regards,

David

[Amit Nex]

@David Boot-Olazabal - Thanks David. The Spotfire documentation for the Action Log Categories says that it cannot be used in case the navigation takes place using custom action control on the text area. I am bit confused. 

[David Boot-Olazabal]

Hi,

I understand why you are a bit confused.
Not sure if this got fixed in the latest version (I'm on Spotfire 14.4 HF4), but you can see that I have been able to capture the 'clicks' in my Actionlog table (the set_page LOG_ACTION is accompanied by the ActionLink value in column ARG6):

My dashboard looks like this (first tab):

Second tab:

And repeating the same actions (clicking on the action control buttons), adds an extra set of rows in my Actionlog table (row number 3 and 4).
I also manually clicked on the tabs, to give you an idea of what the information looks like, compared to the ActionLink value in ARG6 (it then says PageTabs; row number 1 and 2):

So I would encourage you to test this out on your side, and see if the entries pop up in your Actionlog table as well.

Kind regards,

David

[Amit Nex]

Hi @David Boot-Olazabal - Thanks very much. I do see the information I am looking for. This is very helpful.

[Niklas Amberntsson]

@Amit Nex and @David Boot-Olazabal Thank you both for noticing the issue in the documentation. I can confirm that the set_page event is logged when using action controls and also when switching pages using Iron Python scripts. I will make sure the documentation is updated.

[Amit Nex]

Hi @David Boot-Olazabal @Niklas Amberntsson - I have another question regarding the User Activity Log. I am trying to capture the time duration a dashboard has been accessed at the user level. I see 2 tables "AUTH_LOGIN" and "AUTH_LOGOUT" with the login and the logout information at the user level. I tried mapping the tables with the "ACTION_LOG" using the username field. However, this gives me the same login and logout information for all the dashboards accessed by a user.  Thanks for your help.

Edited October 4 by Amit Nex

[David Boot-Olazabal]

Hi Amit,

If I'm not mistaken, the auth_login and auth_logout tables capture information for the session as whole.
The actionlog table should hold the information about individual analysis files. But it's quite hard to make that connection indeed.

Let me have a look at this request and see if I have can find anything you can work with/start from.
I'll come back to you as soon as I have something useful.

Kind regards,

David

[David Boot-Olazabal]

Hi Amit,

I think the closest you'll be able to find information on this, is indeed the actionlog.
I assume your interested in the time a web user has spend on an analysis file.

This information is logged, but you need to do some calculations before you can see the time spend.
The initial information is logged as per the below screenshot:

The calculation you need to create, is between the logged_time or original_time of the lines load (or load_start) and close. The ARG columns, with the exception of ARG1 being AnalysisDxp, don't hold information that could help.

One way to achieve this, is to create a new data table that has been pivoted (add transformation).
When you pass ID2 and ARG2 as row identifiers, the Log_Action as columns and the original_time as value columns, you would end up with something like this.
In this cross tab, you can then create a datediff calculation that calculates the difference between load_start time and close time:

The calculation is as follows (duration being the one that does the subtraction of the load_start and close dates:
UniqueConcatenate([load_start]) as [load_start], UniqueConcatenate([close]) as [close], UniqueConcatenate(DateDiff([close],[load_start])) as duration

Kind regards,

David