NLP- Sentiment Processing for Junk Data takes time

问题

I am trying to find the Sentiment for the input text. This test is a junk sentence and when I tried to find the Sentiment the Annotation to parse the sentence is taking around 30 seconds. For normal text it takes less than a second. If i need to process around millions of data it will add up the time to process. Any solution to this.

String text = "Nm n n 4 n n bkj nun4hmnun Onn njnb hm5bn nm55m nbbh n mnrrnut but n rym4n nbn 4nn65 m nun m n nn nun 4nm 5 gm n my b bb b b rtmrt55tmmm5tttn b b bb g bn nn n h r ret n nun bn d. B bbbbbbbbbbr bung NHnhn nn nk, v v v n gain t g 4gnyhimmigration ndn nb NVnb bin uny 7 nbbbbbnn vv bbvb ninn njnj n4 nm n km n n n cb j bun. Nhfnt bn nn. N hm nn nun m bum my b mmmnbjk nn n by nn nun nun n nun nn bn n nhn n nn n n m NH nb4mnm mkn 4 n n n n hm r b rnfngg4d in b nut mmmkmmm5 bbjn n n ij BBM 8u8i by nun n.nn hm n. n4n By 4n4n bunny RN bny hm j mi. Nymmn FBT not mn n n nm g by n n nnm? Rnyb vCard n5 Yu nn n n n n nt .nm mn nt n nb n n n n by y5nnnhyyh h b b nt njj n m f4n re";
Properties props = new Properties();
            props.setProperty("annotators","tokenize, ssplit, pos,parse,sentiment");
            StanfordCoreNLP pipeline = new StanfordCoreNLP(props);
            Annotation annotation = pipeline.process(text);

For the NLP sentiment calulation I have to parse the data and i think thats the reason its taking time.

thanks

回答1:

Yes, the standard PCFG parser (the one that is run by default without any other options specified) will choke on this sort of long nonsense data. You might have better luck using the shift-reduce constituency parser, which is substantially faster than the PCFG and nearly as accurate.

来源：https://stackoverflow.com/questions/30413885/nlp-sentiment-processing-for-junk-data-takes-time