limitation-recognizer

A program to recognize self-acknowledged limitation sentences in biomedical articles

The repository contains the source code for the system described in the article Automatic recognition of self-acknowledged limitations in clinical research literature. The best performing rule-based system is presented (gov.nih.nlm.limitations.RuleBasedLimitationSentenceRecognizer), as well as the rule-based baseline (gov.nih.nlm.limitations.RuleBasedLimitationSentenceRecognizerBaseline).

Usage

To replicate the results, run gov.nih.nlm.limitations.RuleBasedLimitationSentenceRecognizer with three arguments:

• DATA/XML: directory that contains the parsed XML of the test set
• DATA/limitation_sentences_final.txt: gold annotations
• Output file name (after the run, this file should match DATA/rule_based_test.out.txt)

The parsed XML is generated from PubMed Central XML using gov.nih.nlm.limitations.CorpusParser.

Processing plain text files

To process articles in plain text, run gov.nih.nlm.limitations.CombinedPreprintLimitationRecognizer with two arguments:

• Input directory: a directory of plain text files
• Output file: the file for output (output is in JSON format)

Note on Stanford CoreNLP package

Stanford CoreNLP model jar file that is needed for processing raw text for lexical and syntactic information (stanford-corenlp-3.3.1-models.jar) is not included with the distribution due to its size. It can be downloaded from http://stanfordnlp.github.io/CoreNLP/ and copied to lib directory.

Contact

• Halil Kilicoglu: (halil (at) illinois.edu)