IBM Watson™ Discovery Service Ideas

We've moved...

You'll be redirected shortly, we've moved to our new idea portal:

Display correct document names in the SDU interface.

Nokia is using WDS for their HR Chatbot, and trying out long tail search optimisation through SDU. Anyhow, in the SDU UI's dropdown file list, the files uploaded from the CAESAR/SEA (chatbot asset) Admin Panel to Discovery always get "_" as the extracted_metadata.filename. The asset team researched other possibilities to submit a document to Discovery via the Discovery API but there is no other way than how it's done now (not in document format - submitting the filename as part of the request parameters but Discovery does not seem to correctly interpret it).

Ideally WDS could correctly interpret the metadata and include the correct extracted_metadata.filename. One option would be having the UI populate the document names from the FileName field, but obviously that may have impacts at other clients.

  • Guest
  • Dec 8 2019
  • Already exists
Why is it useful?
Who would benefit from this IDEA? As a Nokia (customer) chatbot project team member, I am able to optimise the long tail search results with the SDU, so that business value can be realised and the use of the chatbot extended within the organization.
How should it work?
Idea Priority
Priority Justification
Customer Name
Submitting Organization
Submitter Tags
  • Attach files
  • Admin
    Christophe Guittet commented
    December 09, 2019 13:37


    The filename is already extracted when you are uploading a file through API. Did you try uploading your files with the following code example (NodeJS Request) or equivalent in another language?
    var fs = require("fs");

    var request = require("request");

    var options = { method: 'POST',

    url: '[environment_id]/collections/[collection_id]/documents',

    qs: { version: '2019-04-30' },


    { 'cache-control': 'no-cache',

    Authorization: 'Basic xxxxxxxxx',

    'Content-Type': 'application/json' },


    { file:

    { value: 'fs.createReadStream("/C:/path/to/your/file.pdf")',


    { filename: 'file.pdf'} },

    metadata: '{"url":""}' } };

    request(options, function (error, response, body) {

    if (error) throw new Error(error);