For integrating into Solr, I've used Word2vec to improve rankings the synonyms it finds (boosting synonyms by how similar they are to the query). In English Word2vec tends to think plurals are synonyms.
I did a talk on this with more details: https://www.slideshare.net/GarySieling/ai-with-the-best-buil...
Sense2Vec might also help: https://github.com/explosion/sense2vec
There is also a project called Vespa, which looks potentially interesting as a replacement for Lucene - https://github.com/vespa-engine/vespa