5 ms·
Parsing enumeration and chronology data for serials. E.g. "v.1" is obviously volume:1. But throw in years, parts, editions, copies, supplements, numbers, page r
by ahi 7y ago
Parsing enumeration and chronology data for serials. E.g. "v.1" is obviously volume:1. But throw in years, parts, editions, copies, supplements, numbers, page ranges, etc, shit gets weird.
- crossroads091 7y agoYou might find some parallels to the field of citation field extraction - where citation/bibliography strings are parsed into author, affiliation, conference name, page/volume of a journal the article appeared in, etc. It is a well-studied problem in NLP and CRF-based taggers have known to yield the best results.