Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Willem Melder
    @wmelder
    Pretty white!
    roelandordelman
    @roelandordelman
    The 'data sources' page has too much text. We should think about a better user design here where we move the information into documentation and faqs. Make clear that the purpose of this view is to browse data collections that are in the MS via CKAN
    lilimelgar
    @lilimelgar
    Yes, indeed. However, something that we would like to do, if you agree (after discussion during last WP5 meeting) is to seamlessly integrate the knowledge base part into the MS menu, as well as the blog (and make the URLs transparent, to avoid having two websites). This was Lora's suggestion. I discussed this with Frank Strater and he gave some ideas on how to do this, which are not yet discussed with you and Jaap. I will show you the two options soon.
    Mark Vallinga
    @MarkVallinga_twitter
    Ik heb even een vraag over de collectie van Beeld en Geluid. Ik heb er al met Eva over gesproken, maar dit was volgens haar typisch iets om aan te kaarten in Gitter. Voor mijn pilot-project ben ik voornamelijk geïnteresseerd in uitzendingen van Sport en Beeld (later: Studio Sport), maar een blik op de resultatenlijst in de Media Suite leert dat er erg weinig complete uitzendingen te bekijken zijn. Het video-icoontje ontbreekt namelijk bij het overgrote deel van de resultaten. De andere uitzendingen lijken niet compleet beschikbaar te zijn (of überhaupt opgeslagen te zijn in de database), maar via de catalogus van Beeld en Geluid ben ik er achter gekomen dat de losse items wel deel uitmaken van de collectie. Die zijn echter alleen niet - voor zover ik weet - beschikbaar in de Media Suite. Is het mogelijk dat ik, om het beeldmateriaal toch te kunnen bekijken, toegang krijg tot extranet?
    Jaap Blom
    @jblom
    Hoi, ah ja Eva kaartte deze kwestie even aan en het is uiteraard een uitzoekwerkje waarom dit zo is. Het moet hier even ingepland worden. Ik gok dat hier volgende week vast iemand naar kan kijken.
    lilimelgar
    @lilimelgar
    An update from Mark: About the problems with viewing the broadcastings of Sport in Beeld: the problem has not been solved yet. When I search in the Media Suite for Sport in Beeld in almost all the cases there isn't a video-icon and it is not possible to view the broadcasting. For example: http://mediasuite.clariah.nl/recipe/default-item-details?id=215557@program&cid=nisv-catalogue-aggr&st=%22sport%20in%20beeld%22 and http://mediasuite.clariah.nl/recipe/default-item-details?id=215532@program&cid=nisv-catalogue-aggr&st=%22sport%20in%20beeld%22. Apparently it has to do with the fact that those programs aren't archived as a whole, but are cut/divided into the individual items of that particular broadcasting. Problematic: those items are not available in the Media Suite. Eva knows about this problem. I think she can tell more about it and what could be the solution (maybe access to Extranet?).
    This has been updated in Github issue #48
    lilimelgar
    @lilimelgar
    Hi @wmelder: as we commented, you will please look into #173 to see if it is a harvesting issue
    lilimelgar
    @lilimelgar
    Hello @MarkVallinga_twitter , we have checked the issue with Sport in Beeld that you reported, it seems those programs are not digitized, we are also checking what is the issue with some missing segments: http://mediasuite-test.rdlabs.beeldengeluid.nl/. We will inform you about this when it is solved, from now, please take note of all issues and report them here ;)
    Berrie van der Molen
    @berrie89_twitter

    I haven't used Gitter so far, so I hope this is the right place to post my observations about the KB dataset that is currently accessible in the MS Single Search test version. Feel free to move this to a more appropriate location.

    I am excited that the KB newspaper dataset is being taken on board now! I suppose actually searching this dataset using a key word isn't working yet? If I enter a query (e.g. "Drugs"), MS takes me back to the screen where I can select a collection (in Single Collection Search).

    • I notice that there is no preview of the articles within MS, is this how it will remain? I know that the PDFs won't be imported due to storage capacity, but could the PDFs be loaded within the MS interface from the Delpher source? [e.g. In Translantis tool Texcavator, the preview image of the article functions as a clickable "link" to Delpher itself]. I often quickly scan the source article, for instance to make sure there are no OCR issues.
    • I don't see the option to generate word clouds made from words in the OCR data (as was possible in AVResearcherXL) - this is an important functionality for the DReAM project. Perhaps this will be only possible in comparative search?
      • When I click on a result, I don't see the national/regional distinction or the name of the newspaper.
    • Suggestion: The description field seems to be used for OCR text, should this field be labeled "OCR"?
    • I notice that not all entries have a date or description field. Is that "in the data"? In that case, the Collection Inspector should allow me to take only the part of the dataset with complete date and description metadata fields to the other recipes (a requirement of the DReAM project already)
    • There are two different ways (metadata fields?) to select results based on date. It should be clear what the difference is between these two options.
    • I assume that the import is not complete yet? I notice that especially after 1951 there is only a small part of the Delpher data in MS (27-11-2017).
    • Minor observation: after updating the date range, you have to press return. An on-screen button to refresh would be more logical in my opinion, because I am always scared pressing buttons may have other (unanticipated) effects on the results.

    I hope I was able to explain these observations clearly, thanks for reading them.

    lilimelgar
    @lilimelgar
    Hi @berrie89_twitter Great input and good use of Gitter :) We are checking, and will let you know soon on how we are going to take care of these issues. Thanks!
    SusanHogervorst
    @SusanHogervorst
    Dear Liliana et al., I was assigned a to do - https://3.basecamp.com/3727471/buckets/3432203/todos/794915086The system is not allowed to search through this collection
    Oops, too soon. Well, to continue: I started to test the Single search recipe, specifically the aggregated index of OH collections. I started to select a collection randomly, in this case The long shadow of Sobibor, and typed a search word (vrouw), but I was notified that 'The system is not allowed to search through this collection', see http://mediasuite-test.rdlabs.beeldengeluid.nl/recipe/single-search?fr=0&sz=20&cids=the-long-shadow-of-sobibor
    Also when I selected another collection, Akkers van Margraten or Getuigen van Theresienstadt, I was sent back to the search page (without any notification this time). So, something is not yet working properly
    SusanHogervorst
    @SusanHogervorst
    By the way: I noticed that the 'meta' level of collection titles, such as Erfgoed van de oorlog en Getuigenverhalen, are no longer there. That is a good thing (because now there are only the titles of the interview projects that were part of these larger 'programs'). Please let me know when Norah and I can test this again! Gr. Susan.
    NorahKarrouche
    @NorahKarrouche
    I had the same problem 'not allowed to search through this collection' :worried:
    SusanHogervorst
    @SusanHogervorst
    today, i tried again, but it did not work. A popup appears saying "loading collection list", and nothing happens after that.
    lilimelgar
    @lilimelgar
    Thanks for testing Susan, I'll show today how it works. Also, as I commented on Basecamp, you need to use VPN now for all collections in the Test version.
    lilimelgar
    @lilimelgar
    Hi @berrie89_twitter We are currently working on improving the import of the KB dataset (see issue #188), we have taken your observations into account. The only suggestion we cannot adopt is including a preview of the articles within the Media suite. This is not allowed because of copyright issues, so we redirect you to Delpher from the MS. More updates will be added to Github (if you follow issue #188). Thanks!
    Julia Noordegraaf
    @jjnoordegraaf_twitter
    Hi @lilimelgar, I'm having the same problem Susan had in December: when opening the 'Add collection' button the list of available collections does not show. Anyone knows why?
    lilimelgar
    @lilimelgar
    Hi Julia. If I remember correctly, the problem in December was related to the test version, which still required the use of VPN. However, since the launching of Version 2 this VPN is not required anymore (at least not to access the list of collections). Could you please try to clean your cache in the browser you are using before checking further? thank you
    Julia Noordegraaf
    @jjnoordegraaf_twitter
    Hi Liliana, indeed, cleaning the cache did the job, sorry !
    Julia Noordegraaf
    @jjnoordegraaf_twitter
    It says "Geen video met ondersteunde indeling en MIME-type gevonden"; anyone?
    NorahKarrouche
    @NorahKarrouche
    @wmelder how long have the OH collections from Verhalis been part of our registry? I only noticed them now. Also, would it be correct, based on the registry, to state that 71 collections are now (to some extent) useable and searchable for researchers based in NL?
    NorahKarrouche
    @NorahKarrouche
    @wmelder issue solved.
    Mark Vallinga
    @MarkVallinga_twitter
    Hi everyone, I have a question about the available material in the Media Suite. I was searching for broadcastings of sport programs, historical football matches and other sport related content in the Media Suite, but there seems to be a problem with the material. Earlier, I already found out that most of the broadcastings of Sport in Beeld weren't available (or more specifically playable) in the Media Suite. There is a detailed description about each specic item available in the catalogue of Beeld en Geluid, but the material hasn't been 'transferred' to the Media Suite. So that means I can't watch it, which makes it hard for me to compare the reporting in the newspapers with how it is broadcasted on television. Unfortunately, that problem hasn't been solved yet. Is there more news about that issue? But it looks like there are more programs about sport missing in the Media Suite. For example: in 1962 the Dutch football team played matches against Belgium (14-10) and Switserland (11-11). When I search in the Media Suite for these specific games (and broadcastings of Sport in Beeld around these dates) no results appear, but when search in the catalogue of Beeld en Geluid I find the broadcastings of Sport in Beeld and Polygoon. And also in 1966: on 30-11 the Netherlands played a match against Denmark. Again no results in the Media Suite, but in catalogue of Beeld en Geluid itself I see the whole match is digitized and a short version is broadcasted in het Weekjournaal. How is that possible? Is it maybe a copyright-issue or something like that? Gr. Mark
    lilimelgar
    @lilimelgar
    Hi Mark, good news, the development team has been working on this and they found the problem, and a solution! You will be able to find some of these materials at the Summer school, and before when version 3 is ready for testing . You will hear from us when that happens. @mwigham: I thought you would like to see Mark's question (he is one of the CLARIAH research pilot scholars). Thanks!!
    Christian Gosvig
    @chrisgolesen_twitter
    Hi, Eef, Bregt and I have two questions concerning Desmet data. 1. In the inspector tool, both for the paper and film collections all metadata fields always display as 100 percent complete. 2. when retrieving an OCRed document from the paper collection the search term used is not highlighted in the metadata (nor the document) and can also not be found by using cmd f (both in metadata and doc). It would be very nice to be able to do that. Thanks! Chris
    lilimelgar
    @lilimelgar
    We also would like to have a reindexing of the Desmet papers to distinguish (add a label) to the fields/files that are enrichments, to distinguish it from the base (excel fields)
    NorahKarrouche
    @NorahKarrouche
    Not sure if I am in the right room for this, but I just found out that a search with the 'explore' tool also generates OH data, just not OH data stored at DANS. I understand why this might be the case, but a new user does not get why explore and search give different results in the MS.
    mwigham
    @mwigham
    A lot of data at NISV is described but not yet digitised, it would be good if more can be digitised
    lilimelgar
    @lilimelgar
    @MarkVallinga_twitter Hi Mark, we are curious to know if you can find in the MS v3 the programs you were looking for. @mwigham has worked in solving the problem you described here on May 7. Best, L.
    lilimelgar
    @lilimelgar
    @chrisgolesen_twitter Hi Christian, thanks for your input, and apologies for slow reaction. About "1. In the inspector tool, both for the paper and film collections all metadata fields always display as 100 percent complete." This is work in progress, partially solved, and will be ready/fixed in the next version (issue #471). The second issue is also going to be taken into account in the redesign of the resource viewer.
    lilimelgar
    @lilimelgar
    Not sure if I am in the right room for this, but I just found out that a search with the 'explore' tool also generates OH data, just not OH data stored at DANS. I understand why this might be the case, but a new user does not get why explore and search give different results in the MS.
    Thank you @NorahKarrouche , this room is fine :). The Explore tool (DIVE) is currently integrated only partially. Some of the data/collections in the DIVE tool come from different indexes than those used by the Media Suite. At the summer school, we focused on the only collection that is the same (shared) among both "tools", which was Open Beelden. But a full integration of the data is current work in progress in the context of the Linked data work (priority for the next phases of CLARIAH). I will add this to the documentation to avoid confusion. Thanks!
    A lot of data at NISV is described but not yet digitised, it would be good if more can be digitised
    @mwigham That would be great... but because that is a bit out of the scope of CLARIAH, it would be interesting to integrate this information in the metadata somehow, that is, when the user finds a resource that is not digitized, that this was made explicit somewhere in the metadata, to avoid confusion, for instance, if an item doesn't play/show for other reasons. At the moment there is an automatic detection: if the resource has a media, the resource shows an open icon in the resource viewer, if there is no media, the eye is closed. But this is not in the metadata. Would this be possible to integrate somehow, at least for B&G?
    lilimelgar
    @lilimelgar
    @NorahKarrouche Hi Norah, hope you are well. We are polishing the icons in the Media Suite that detect the media type, since now, they all show video for the oral history collections. From your experience in using this collection, do you have examples of an audio-only interview in the Media Suite?
    NorahKarrouche
    @NorahKarrouche
    @lilimelgar @mwigham The updated CKAN info looks good! About "All the DANS oral history "thematic collections" have been harvested.'' That is not the case. As I said before, I am drawing up a new list of OH in the MS, and other collections in DANS and elsewhere that we should harvest. I would also add a note about the difference between projects and thematic collections (I can do that). Would it be useful if I added a bit more on how to use oral history?
    NorahKarrouche
    @NorahKarrouche
    @lilimelgar I no longer see the videos themselves, so I can't check. It could be that icons that signal visual actually are audio only, right?
    lilimelgar
    @lilimelgar
    @lilimelgar I no longer see the videos themselves, so I can't check. It could be that icons that signal visual actually are audio only, right?
    Hi @NorahKarrouche , thanks for letting us know. Could you please send us a screenshot of the issue you are having with the videos? Is it perhaps related to the issue described here? http://mediasuite.clariah.nl/documentation/troubleshooting
    About your first question (CKAN pages), we are looking into it, and we will reply later during this week
    jjnoordegraaf
    @jjnoordegraaf
    Hi @lilimelgar and @mwigham, While exploring the collection visualizations in the CKAN repository for a chapter I am writing I realized we only have those for the NIBG and EYE collections. Is there a specific reason for this? Or can we provide these visualizations for any of the collections registered in CKAN?
    lilimelgar
    @lilimelgar
    Hi @jjnoordegraaf, Mari also made visualizations for the aggregated collection made of DANS OH collections, here: http://mediasuitedata.clariah.nl/dataset/dans-oral-history, and for Meertens' Sound Bites: http://mediasuitedata.clariah.nl/dataset/soundbites. Also for the KB newspapers: http://mediasuitedata.clariah.nl/dataset/kb-newspapers-test. When Mari and I discussed how to do this, we thought it wouldn't make much sense to add it to all the individual OH collections, thus, we have them for the aggregated version only. We also decided not to do it for Open Beelden, since this is a sort of very specific sub-set of NISV and Eye, but we could also do it later. Please let us know if you have suggestions for improving. Best, L. (ps. Mari is on holiday at the moment, but @mwigham: please feel free to add any additional clarification when you are back if I missed something)
    jjnoordegraaf
    @jjnoordegraaf
    Many thanks @lilimelgar, I will check it out. The question whether only aggregated visualizations for the OH collections make sense, is an issue I would like @NorahKarrouche to reflect on; in each case useful to include a sentence for each separate collection that refers to the aggregated visualization. Thanks!
    lilimelgar
    @lilimelgar
    Hi Julia, thank you for the suggestion to include a note on each collection, I will check this with @jwassenaarb (to do it automatically, since there are more than 60 now) and with Norah (the text). I can also note that @NorahKarrouche has been involved in the steps that led to create the aggregated collections together with the development team, and that @mwigham has worked with her on the visualizations :). Happily, oral historians will have the chance to get an introduction to this work at the forthcoming OH workshop :).
    mwigham
    @mwigham
    Visualisations can indeed be created for any collection in CKAN, the bottleneck is in putting them on the website, as that currently involves a manual step. As Liliana mentions, it is also always important to consider whether a given visualisation helps the scholars or not. That's where your feedback is so useful.