We present a defense platform against text-based message spam on SmartPhones. We focus in particular on Short Message Service (SMS) based SPAM. Our solution relies on a social network based collaborative approach to filter this type of spam using Bloom filters and content hashing. We detail the design of the supporting framework and validate its efficiency in minimizing false positives and limiting the storage space. We show that a content hashing based approach provides better lookup than a bloom filter with the same storage space and false positive probability close to $10^{-9}$.