Using machine-readable vocabulary like schema.org is generally a good way to allow people and non-humans to re-use the information you supply on the web. For example, this will make it easier to process your publication data and get a formatted, database-like normalized list out of it.
One might think that search engines use schema.org vocabulary to analyze your website and hence make it better visible on the web. However, what search engines do is somewhat a black box (in the sense that something is happening in there, generating output from user-driven input, but the algorithm inside is not known in detail).
If you would like to reach a better visibility on the web for your publications, you might also like to consider using ORCID which is based on linked-data principles, widely known and used in library and information services - and some research funding institutions are starting to require complete ORCID profiles to be submitted with proposals anyway.