Is anyone applying modern machine learning models to the undeciphered Indus script? Seems like an ideal problem for pattern recognition to solve.
So, this isn't that useful, for deciphering the Harappan script, but I realized that since Sumer (most likely) traded with IVC, that means they must have some loanwords. I looked up Sumerian trade routes with Meluhha (IVC), and I found they mentioned tin a lot, which is supposedly spelled AN-NA, but pronounced nag-ga /nak.ka/, so I figured that must be a loan word. After looking up the dravidian words for it, I reconstructed the word *nakkara or *takkara, which would have meant tin. Is this something new or was it already known? Also, does anyone disagree or think I'm wrong? Lastly, I think the name for their civilization is *melakham /melaxam/ or *melakam /melakam/ because Asko Parpola identified Meluhha as coming from *mel (elevated) and *akam (surface), that one isn't entirely my discovery though.
Does anyone know if any effort is going on to decipher IVC script? Where can I find information on latest developments?