Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Iris van Vliet
    @irisvvliet_twitter
    Me and Marijn have the following comments on the comparative search:
    Interface:
    • The interface of the tool does not invite to use it. For example, with the earlier version of AVResearcherXL and the current version of DIVE, you immediately see a search bar which invites you to type and search something, and based on that you can find out how the tool works.
    • There is a bug in the time slider which shows weird date values. With the query ‘Koningin AND Beatrix’ the timeline within the Beeld en Geluid collection runs from 1919 till 2061. With the query ‘Nederland’ it runs from 1 to 2061. When you leave the search bar empty and click ‘submit’ anyway, the time slider even shows a period from the year 1 to 2999. [See printscreens below]. This is also an issue with the IPNV collection (2006-2042). For all the other collections, the timeslider works as it should.
    Iris van Vliet
    @irisvvliet_twitter
    image.png
    image.png
    image.png
    Iris van Vliet
    @irisvvliet_twitter
    • When you search for two queries at the same time, you do not see the results next to each other, but below each other in a semi-continuous overview. This makes it kind of chaotic, since it is unclear where the results from query 1 stop and where the results from query 2 start. I would personally prefer the two result columns to be shown next to each other (like in AVResearcherXL) or otherwise clearly divided.
    • I like how you can choose which metadata you want to be shown in the tables underneath the query. However, the names of the metadata fields should be renamed, since it is not always clear what information the different fields contain (I believe this is already on our to do-list). Also, I expect that not every metadata field has been filled in every case. For that reason, the future manual should include a disclaimer that not every metadata field is complete for all the results and that the shown tables are not always a representative result.
    • It would be useful to see the results in the timeline in both a comparative and an absolute scale (like in AVResearcherXL).
    • It would be useful to be able to click on a year in the timeline to see the results of that specific year for further analysis.

    Collections:

    • After ‘adding a query’, you’re asked to select a collection. However, a (new) user might not know what is in these collections. It would be a good idea to add short descriptions to the collection-boxes, about the content of that specific collection. For example: it is somewhat confusing that there are multiple Beeld en Geluid collections to search through (either the video results, the audio results, tv-guides, or a total collection that contains all of these), but it is unclear that when you only want to search in either one of the categories, you need to select another collection.
    • This could be ‘solved’ by visually placing the three collections next to each other when selecting a collection (sorting them by institution instead of collection name).

    Information/user friendliness:

    • While searching, I miss information on how the tool works. I would like to have a short user manual on the tool, which tells me more about how it works, what the different options do, etc. (Like the 'Help' page in AVResearcherXL) --> we already discussed this in the meeting on May 2nd. To be continued...

    Results:

    • If you click on a result and then on the ‘back’-button (previous) in your browser (I used Google Chrome), all your results and queries have disappeared. The interface does not offer a back/previous button. This is very frustrating!!!
    • You can also see ‘all data’ after clicking on a result. However, this data is shown in a programming language/jargon. The average humanities scholar probably finds this confusing, so it should be visualized differently (also on our to do list!)
    Ofcourse there is some overlap in what we already discussed earlier, but we thought we'd just throw anything in here. More to follow on friday!
    Let me know if more print screens are needed to clarify our comments!
    lilimelgar
    @lilimelgar

    REPLIES.

    Interface and interaction design

    lilimelgar
    @lilimelgar
    I see a number of comments related to the user friendliness of the interface and the interaction. This work is going to be done after July (front-end design mostly). It would be very useful if you add the ideas for improvement to the Design documents. However, keep in mind that AVRXL only works with two queries at the same time, while this recipe works with a bigger number of queries defined by the user, that is why, for instance, presentation of search results cannot be done horizontally. The "back-button" issue was added before: #38.

    User friendliness of metadata schemas

    We are working on a plan for this, there is no clear deadline yet, but it is included in the planning
    lilimelgar
    @lilimelgar

    Time selector

    Please see Github issue #75

    Facet selector

    lilimelgar
    @lilimelgar
    "I like how you can choose which metadata you want to be shown in the tables underneath the query...": This, we consider, is a very advanced and interesting feature for scholars, since this was not possible to do before in other systems (i.e., choosing which facets you want to use to build your query). The improvements have to be done at the levels you indicate: selecting the basic facets that should be common (this requires creating a common schema for all collections), and naming the fields using the labels that are more user friendly (previous issue about "user friendliness of metadata schemas").

    Timeplot

    "It would be useful to see the results in the timeline in both a comparative and an absolute scale (like in AVResearcherXL)." This was already added as Github issue #80
    lilimelgar
    @lilimelgar
    "It would be useful to be able to click on a year in the timeline to see the results of that specific year for further analysis." This was also reported in the Collection inspector (both recipes use the same component: timeplot). See issue #83

    Collection selector

    Hierarchies between datasets and sub-collection snippets are desirable. See Github issues #87 and #50.
    lilimelgar
    @lilimelgar

    Documentation

    In relation to your suggestions about user manuals, please keep in mind that documentation is not ready yet ;). Work in progress (contributions are accepted) in the guidelines documents or in the "documentation_knowledge base" Gitter room here.
    Jasmijn Van Gorp
    @j_vangorp_twitter
    I am using the test version. When I set the end date (in calendar) and run it, I can't change it anymore. The start date is changeable, but the end date isn't
    lilimelgar
    @lilimelgar
    I have tested this (also in test version) and I don't notice the same issue. Could you please record a short screencast to show how it happens? Also, please keep in mind that the date range depends on the available dates (earliest, latest) for each type of date that is selected in the date field
    Jasmijn Van Gorp
    @j_vangorp_twitter
    Hi all, I am wondering which field 'description' is in the search boxes? Is it an aggregation of some other metadata fields? Which ones? Thanks!
    lilimelgar
    @lilimelgar
    Hello Jasmijn, good question, yes, it is an aggregation of several fields, which change depending on the collection. To see which fields are aggregated per collection, you can follow these steps: use Chrome, open the developer tools (three vertical dots in the right top of your screen), and go to "tools" in the MS as usual to load your collection, then click on the "description" field cluster, and observe in the console (network tab) which fields are aggregated. There is a screenshot in this document, point 3 of the exercise, and I made a screen cast (available here). We had implemented a tool-tip showing the composition of the aggregation, but it was taken out to offer the option to add more than one cluster in the query. We will provide more details in the documentation later on. I hope this helps, let me know if there are further questions/suggestions.
    Jasmijn Van Gorp
    @j_vangorp_twitter
    thanks! I figured it out; it is indeed an aggregation