Title: Chinese Lexical Resources for Gender, Number, Animacy 
Authors: Zhiyi Song, Jiahong Yuan, Xiaoyi Ma, Stephanie Strassel


1. Introduction

DARPA's Deep Exploration and Filtering of Text (DEFT) program aimed to
address remaining capability gaps in state-of-the-art natural language
processing technologies related to inference, causal relationships and
anomaly detection. In support of DEFT, LDC provided source data
and core resources for system development.

Gender, number, and animacy are lexical indicators that can be useful in
the detection of person mentions. LDC created the Chinese Lexical Resources
for Gender, Number, Animacy Corpus by extracting information from newswire
texts in the Chinse Gigaword Corpus (LDC2011T13). The corpus includes
dictionaries of Chinese animate nominals and names; Chinese nominals and
name with gender and number predicted; and other dictionaries of Chinese
nominals, names, verbs and pronouns; each dictionary contains frequency
information as well as the features in question. The data was released to
DEFT performers with an eye to improve performance of core NLP capabilities
such as entity tagging and coreference.

2. Contents


./docs/
  README.txt  --  This file

  ChineseLexicalResources_v4.pdf   -- Detailed description of how
                                      lexicons contained in this release 
                                      were created
  designators/			   -- All designator lists used in building 
				      gender and entity lexicons

./data/animacy
  This directory contains dictionaries of Chinese animate nominals and names.
  The files contain listings of extracted nominal or name phrases with 
  frequency of occurence. In each line of the following .lex files, the 
  (possibly multi-word) noun phrase is followed by a tab and then its 
  frequency from the query result.

	appositionPersonName.lex
	nameSubjectBa.lex
	nounSubjectBa.lex
	org_name_NRNN.lex

  In each line of the following .lex files, the (possibly multi-word)
  noun phrase is followed by a tab and then its predicted entity type
  and frequency from the query result. The three entity types that are
  predicted include Person (per), Organization (org) and Location
  (loc), where location covers GeoPolitical Entity, Location and
  Facility.

        named_entity_apposition.lex
        named_entity_NRNN.lex

./data/gender_number
  This directory contains dictionaries of Chinese nominals and names
  with gender and number predicted (refer to
  ChineseLexicalResources_v4.pdf for details of queries). The files
  contain listings of extracted nominal or name phrases and their
  possible gender counts (as predicted based on surrounding
  text). 

  In each line, the (possibly multi-word) noun phrase is
  followed by a tab and then columns holding the counts for the
  corresponding gender/number. 

  Gender and number are coded as: 
       MS    male singular 
       FS    female singular 
       NP    neutral plural 
       IS    inanimate singular
       NS    neutral singular 
       IP    inanimate plural 
       NS    neutral singular

      conjunctive_possessive_name.lex
      conjunctive_possessive_noun.lex
      nominative_predicate_name.lex
      nominative_predicate_noun.lex
      verb_nominative_name.lex
      verb_nominative_noun.lex

 The following two dictionaries include person names, each of which is
 singular.  The gender is coded as:

	F	female
	M	male
	N	neutral

      gender_designator_apposition_name.lex
      gender_designator_NRNN_name.lex

./data/other
  This directory contains dictionaries of Chinese nominals, names, verbs and
  pronouns. The files contain listings of extracted words and their frequency. 
  In each line, the (possibly multi-word) noun phrase is followed by a tab and 
  then its frequency from the query result.
      name.lex
      noun.lex
      pronouns.lex
      verb.lex

3. Acknowledgement

This material is based on research sponsored by Air Force Research
Laboratory and Defense Advance Research Projects Agency under agreement
number FA8750-13-2-0045. The U.S. Government is authorized to reproduce and
distribute reprints for Governmental purposes notwithstanding any copyright
notation thereon. The views and conclusions contained herein are those of
the authors and should not be interpreted as necessarily representing the
official policies or endorsements, either expressed or implied, of Air
Force Research Laboratory and Defense Advanced Research Projects Agency or
the U.S.  Government.

4. Copyright

5. Contact Information

  Zhiyi Song         <zhiyi@ldc.upenn.edu>     Research Project Manager
  Stephanie Strassel <strassel@ldc.upenn.edu>  DEFT PI

---------------
README log
   Created by Zhiyi Song, July 14, 2014
   Updated by Zhiyi Song, July 17, 2014
   Updated by Zhiyi Song, July 21, 2014
   Updated by Zhiyi Song, November 20, 2014
   Updated by Zhiyi Song, February 9, 2015
   Updated by Zhiyi song, February 10, 2015
   Updated by Zhiyi Song, October 7, 2015
   Updated by Zhiyi Song, November 16, 2015
   Updated by Ann Sawyer, January 12, 2017