rodekruis/social-media-listening

Define user segments

Opened this issue · 0 comments

jmargutt commented

Context
Once we have an understanding of the proof of concept with higher quality labels, it will be good to understand for which users will SML be of value.

Proposal
This will be a combined effort from the SML team to understand the segments where we can ship SML, most likely indicated by

  1. Language in scope of the model (now limited to the languages supported by Azure Translator)
  2. Data sources supported by the model: Telegram, Twitter/X, Kobo, ...?

Upon Richard's feedback, also use this to define:

  1. The eligibility criteria for this product (Language, crisis type, region, time to value)
  2. The proof of concept without Argilla