Improving SO-CAL and VADER

[hw016]

Hi,

Thank you for providing this open resource. I wanted to ask if it would be possible to expand the lexicons in sentiment dictionaries like SO-CAL and VADER by considering their custom algorithms. For example, could we expand existing lexicon entries by adding common negators (e.g. "not") or intensifiers (e.g. "very") in front? This could transform entries like "great" into variations such as "not great" or "really great."

I read your documentation on enhancing resources like the Dictionary of Affective Language (DAL) and Norms of Valence, Arousal and Dominance (NoVAD). Since you reflected authors' discussion points to improve these lexicons, I was wondering if similar enhancement techniques could be applied to VADER and SO-CAL as well.

Thanks!

[nick-sh-oh]

Hey @hw016,

Thank you for your request - That's actually a great idea! Currently, I'm in the process of preprocessing Econ/Finance domain-specific dictionaries. I'll make sure to incorporate your recommendations for VADER and SO-CAL once I upload those dictionaries.