Techniques for predicting the spectra of materials using molecular metadata

A system comprising:

• one or more data processors;
• and
• a non-transitory computer readable storage medium containing instructions which, when executed on the one or more data processors, cause the one or more data processors to perform operations comprising:
• inputting, by a computing device, a text string comprising a structural representation of a material to an encoder of a natural language processing (NLP) model implemented in a deep neural network;
• generating, using the encoder of the NLP model, an encoded representation of the text string comprising latent chemical bond information of the material;
• mapping, by the computing device, the encoded representation comprising the latent chemical bond information to a spectrum array, the spectrum array comprising predicted spectroscopic data of the material;
• and
• outputting, by the computing device, the spectrum array.

A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, including instructions configured to cause one or more data processors to perform operations comprising:

• inputting, by a computing device, a text string comprising a structural representation of a material to an encoder of a natural language processing (NLP) model implemented in a deep neural network;
• generating, using the encoder of the NLP model, an encoded representation of the text string comprising latent chemical bond information of the material;
• mapping, by the computing device, the encoded representation comprising the latent chemical bond information to a spectrum array, the spectrum array comprising predicted spectroscopic data of the material;
• and
• outputting, by the computing device, the spectrum array.