Open danydvd opened 3 years ago
Very cool.
It would probably take a day or so to add a subject page like the one you point to.
Thanks @danydvd Can you clarify for James if and how this new data should be incorporated into the application (e.g., used in search, etc.).
I will add a separate issue for the addition of a page for each subject so we can have it noted. I think it is lower priority at the moment than widget interactivity, but definitely something that would be great to have in the future.
@sfarnel @jchartrand I don't think that we need to modify the existing (i.e. I can't imagine a way that we would use the subjects urls in search or search result page as we are not faceting or visualizing by subjects).
@sfarnel I am not sure if this was the idea that you were referring to?
Thanks @danydvd we do use subjects for the dropdown as well as the word cloud (which would use the labels) so I assumed we might be able to make use of them there.
@sfarnel @danydvd
I'm not totally sure what to do here.
I could hyperlink the subjects that appear on the full record page, e.g.,
http://206.167.181.124/record/e6ef3b77306f69003ece2a41c077840b
so that if you click on a subject it takes you to the corresponding page url from subject_url.
Is that what I should do?
Thanks @jchartrand I think what we have in mind here (@danydvd please chime in as well!) is that since we have indexed the URIs for the subjects, if we could use that data for matching (in search and facet) but of course display to the end user the text strings then we might get even better results? But perhaps we are already doing this or it maybe harder to do than we think?
I totally agree @sfarnel. If it is feasible that we use indexed URIs (subject_url in SOLR) for search/facet and string value (subject field in SOLR) for display (e.g. http://206.167.181.124:8000/getitem/http://canlink.library.ualberta.ca/thesis/8f2ad2192ac56a61655ecb583a4cd03f)
@sfarnel @danydvd
I'm for whatever reason really having a hard time with this one. I guess I don't see how the url versions of the subjects will change matches?
Here is a sample record:
{
"id":"http://canlink.library.ualberta.ca/thesis/fb3942f8d9a6f063504ce5a587b6c8a7",
"year":[2003],
"title":["Le développement de la personnalité de l'homme de l'adolescence au milieu de la vie : approches centrées sur les variables et sur les personnes"],
"institution":["http://canlink.library.ualberta.ca/institution/Université_de_Montréal"],
"institution_name":["Université de Montréal"],
"institution_short":["U de Montréal"],
"creator_url":["http://canlink.library.ualberta.ca/person/8ac3c953a43e460d2951dc639cdca8ca"],
"creator":["Morizot Julien"],
"lang":["http://id.loc.gov/vocabulary/languages/fre"],
"link":["http://search.proquest.com/docview/305239136?accountid=12543"],
"creator_first":["Julien"],
"creator_last":["Morizot"],
"degree":["PhD"],
"degree_name":["Doctor of Philosophy"],
"subject":["comportement antisocial",
"judiciarisé",
"maturation",
"typologie",
"trait",
"continuité"],
"subject_url":["http://canlink.library.ualberta.ca/subject/433f03787dc72896739f9b371c86cd09",
"http://canlink.library.ualberta.ca/subject/ee3ce66373ed3b716f25889ffa085445",
"http://canlink.library.ualberta.ca/subject/b986b2bf13bf3f2657939047c64cec29",
"http://canlink.library.ualberta.ca/subject/56aa7f225d81ca1bfc651bbf358c4c8e",
"http://canlink.library.ualberta.ca/subject/94a223decbd8cae91c432ca80c2be1f7",
"http://canlink.library.ualberta.ca/subject/3442ee7de4438cffaf8c5ea9b3a05f7e"],
"_version_":1679131393736572928}
For that record, how would matching on:
http://canlink.library.ualberta.ca/subject/433f03787dc72896739f9b371c86cd09
instead of on the corresponding term:
comportement antisocial
change search results? I feel like I'm just not understanding some piece of all this.
Thanks @jchartrand I think we may need to do more work on the data end to enable the kind of functionality we had in mind.
More data work needed; keeping for blue sky purposes
@jchartrand the newest SOLR index (CanLink-new-6) now includes subject_url fields. Since this field is multi valued (values are in a list) the subject_url is also a list and each item in this list corresponds to the item in the subject list.
"subject":["roman", "inceste", "famille", "loi", "nature", "xviiie siècle"],
"subject_url":["http://canlink.library.ualberta.ca/subject/b179a9ec0777eae19382c14319872e1b", "http://canlink.library.ualberta.ca/subject/00e0868463b1ddc2ba31ec9ef835bc35", "http://canlink.library.ualberta.ca/subject/ed81c0311eb5364a8d6d1d1b63f3c51d", "http://canlink.library.ualberta.ca/subject/84ab36b2995bb3949db34038a2b24c64", "http://canlink.library.ualberta.ca/subject/405aaff66082ffe7231d7c1f79926c17", "http://canlink.library.ualberta.ca/subject/be5ade92ef8c916908660a58e4d36ce0"]
@sfarnel @jchartrand @CarlsoFiorention what are your thoughts on having a separate page for individual subjects?(e.g. http://206.167.181.124:8000/getitem/http://canlink.library.ualberta.ca/subject/e52aafb06de949864871fd8885da9b04)