For an ongoing project, we require a large dataset of 1 million+ SMS messages. Messages must be:
Not obscene / bad language
Preferably from live SMS system, so they match all network heuristics
Not full of hashtags
Not breaching anyone's privacy, so cleaned of secure data
Please quote per block of 50,000 messages.
We require a sample of around 100-1000 typical messages to estimate quality before purchase.
We will randomly sample messages purchased to ensure they meet the above criteria