Enhancing Thesaurus or Dédalo Functionalities to Filter Specific Content

Gerard_La Foneria

I’d like to explain the problem we are facing, in order to help you understand a proposal for evolving the thesaurus or Dédalo.
In a project where we use a general thesaurus, we would like to display online the content related to the repression of women, separately from the general thesaurus—essentially as a filter. However, we do not want to create a separate thesaurus or a new branch within it. Our intention is to continue using the same terms from the general thesaurus in order to avoid double indexing.
We have considered two possible solutions, but neither one seems to be viable:
First: In cases where keywords clearly refer to the specific repression of women (such as head shaving), we could implement web programming to display only those terms online. However, for broader terms that do not specify gender—such as extrajudicial killings—this approach would not work, since it wouldn’t distinguish whether the repression targeted men or women.
Second: We could tag the resources themselves (e.g., a document) instead of modifying the thesaurus, which would then allow us to filter accordingly. But we believe that this solution doesn’t apply to interview fragments, because fragments are not treated as standalone resources, and can't be tagged outside thesaurus.
Has anyone encountered a similar situation, or could we think of a solution or an evolution of Dédalo that would support this type of functionality? One idea might be to allow interview fragments to become taggable resources—but when web programming, this could potentially result in too many database calls during thesaurus-based navigation.

Alex

Gerard_La Foneria

Seems a diffusion issue that can be resolved doing an all parents terms field or columns in diffusion ontology of the thesaurus.

In Dédalo you can use the typology/model parallel thesaurus to identify the terms, as toponymy does.

In diffusion context, you can add a column with all parents of the terms (all parents recursive), and filter it in the thesaurus calls. For example:

If you want to get all children of the Tarragonès region (es1_8187) you can make a API to /records with /sql_filter configured with

`parents` LIKE '%es1_8187%'

Something as:

https://my_server_org/dedalo/publication/server_api/v1/json/records?code=My_API_code&db_name=my_database&table=tp_spain&sql_filter=%60parents%60%20LIKE%20'\''%25es1_8187%25'\''&lang=lg-spa

The API will return all toponymies inside the region. Therefore, you can filter other calls with this fixed filter, for example, if you want all municipalities that starting with 'Tarre' in combination with this fixed filter, the API only returns the matches within the Tarragonès toponymy. This way is super fast and you can see an example in the catalog of Moneda Ibérica. It show all children of specific mint inside a full mints, cultures, periods,... thesaurus using the parents way explained.

Finally, you can see a diffusion node in ontology to get all parents here. Just add the add_parents to the v5 properties like this:

{
  "process_dato": "diffusion_sql::map_locator_to_terminoID",
  "process_dato_arguments": {
    "custom_arguments": {
      "add_parents": true
    }
  }
}

I hope is clear.

Best.