lsa - Document Similarity in R -

i new text mining , semantic analysis. trying find similarity of documents. first approach, using jaccard_similarity find documents similar. have read, jaccard_similarity find similar words, perhaps not best approach finding similar documents, thought give try anyway.

the problem facing following. have 200 type documents, , 1000000 b type documents. need see (if any) of documents similar b documents, not need see if type documents similar between them, or if b documents similar between them.

i found "textreuse" package in r perform this. following:

minhash <- minhash_generator(100, seed = 235)  corpus <- textreusecorpus(text=as.character(data$text),                           tokenizer=tokenize_ngrams, n=5,                           minhash_func=minhash)  buckets <- lsh(corpus, bands = 50, progress = false)  candidates <- lsh_candidates(buckets)  scores <- lsh_compare(candidates, corpus, jaccard_similarity,                        progress = true) 

where data$text subsample of complete a+b documents. issue there many documents and, approach, doing unnecessary comparisons (a1 a2 documents, example, or b1 b2).

is there way in faster way? or there proper way semantic analysis?

thank you.