README
This package includes all resources used in paper: "Be Appropriate and Funny: Automatic Entity Morph Encoding", proceeding of ACL 2014.

- character_decomposition
	It contains a Chinese character to radical mapping table that includes 191 radicals (59 of them are characters) and 1328 common characters.

- chinese_character_dict
	It contains semantic interpretation for each Chinese character. Interpretations for characters with multiple pronunciation are included as well.

- NamePair_ch_to_en
	It contains Chinese and English name pairs of FAC(facilities), GPE(geopolitical entity), PER(person) and LOC(location).

- NamePair_en_to_ch
	It contains English and Chinese name pairs from /* todo */.

- negatives_simplified
	It contains negative Chinese words and phrases from "Hownet 情感词(emotional words)"

- positives_simplified
	It contains positive Chinese words and phrases from "Hownet 情感词(emotional words)"

- experiment_results
	It contains human evaluation results for human created and system generated morphs of 9 annotators. Each line is represented as: 
	<entity> <morph> <category index> <true as human created morph and false as system generated morph> <answer for Q1> <answer for Q2> <answer for Q3> <answer for Q4>
	Questions asked in the evaluation:
		Q1. Do you know who this morph refers to?"
			a1. retty sure; a2. Not sure; a3. No clue
		Q2. Did you guess it right?"
			a1. Yes; a2. No;
		Q3. How interesting do you think of the morph?"
			a1. Funny; a2. Somewhat funny; a3. Not funny; a4. Don't know the entity
		Q4. How well do you think the morph capture the characteristic of the entity? (Or does the morph make sense?)"
			a1. Make sense; a2. Make a little sense; a3. Make no sense; a4. Don't know the morph