AplikaceAplikace
Nastavení

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
en:cnk:fictree [2017/12/12 10:44] – michalkrenen:cnk:fictree [2017/12/18 19:24] (current) – [How to cite FicTree] michalkren
Line 1: Line 1:
-===== FicTree - a manually annotated treebank of Czech fiction ===== +===== FicTree: a manually annotated treebank of Czech fiction =====
  
 +The FicTree treebank is a syntactically annotated corpus of Czech fiction. It consists of 135,000 words (166,000 tokens).  The lemmatization, the morphological and syntactic annotation were performed manually.
 <WRAP right 35%> <WRAP right 35%>
 ^ <fs medium>Name</fs> ^^ <fs medium>FicTree</fs> ^ ^ <fs medium>Name</fs> ^^ <fs medium>FicTree</fs> ^
Line 11: Line 11:
 ^ ::: ^ Publication date | 2017  | ^ ::: ^ Publication date | 2017  |
 </WRAP> </WRAP>
-FicTree is a syntactically annotated corpus of Czech fiction. It consists of 12,760 sentences (166,432 tokens).  +===== The composition of the FicTree treebank =====
-The texts come from eight literary works published in the Czech Republic between 1991 and 2007. +
-The text data was manually annotated according to the [[https://ufal.mff.cuni.cz/pdt2.0/doc/manuals/en/a-layer/html/index.html|Prague Dependency Treebank guidelines]] (annotation on the analytical layer). The texts are shuffled into random chunks of maximum 100 words (respecting sentence boundaries).+
  
-=== Annotation procedure === +The FicTree treebank consists of eight literary works published in the Czech Republic between 1991 and 2007. The texts in the treebank include six fiction titles, a children’s fiction  book, and a book of memoirs.  
-The texts were parsed independently by two parsers trained on the Prague Dependency Treebank data (analytical layer). The parsing results were manually  +Most of the texts were first published between 1991 and 2007 except for one text, published in 1969. 
-corrected and the two versions merged. Any differences were resolved manually. +Five texts (80% of all tokens) are original Czech texts, the other three are translations (from German and Slovak).
  
-=== Text details ===+===== The syntactic annotation of the treebank =====
  
-The eight texts in the treebank include six fiction titles, a children’s fiction  book, and a book of memoirs.  +The FicTree treebank was syntactically annotated according to the guidelines for the analytical layer of the Prague Dependency Treebank – PDT ([[http://ufal.mff.cuni.cz/pdt2.0/doc/manuals/cz/a-layer/html/index.html|PDT 2.0]] with revisions [[http://ufal.mff.cuni.cz/pdt2.5/cs/documentation.html|2.5]] and [[http://ufal.mff.cuni.cz/pdt3.0|3.0]]). The corpus was parsed using two parsers: ([[https://sourceforge.net/projects/mstparser/|MST Parser]] and [[http://www.maltparser.org/|MaltParser]]) trained on the PDT a-layer train data. The results were manually corrected by annotators, then merged. Any differences between the two versions were resolved manually by another annotator.
-Most of the texts were first published between 1991 and 2007 except one text, published in 1969. +
-80% of the texts are original Czech texts, 20% are translations (from German and Slovak).+
  
-=== References === +===== Access to the treebank =====
-Tomáš Jelínek, 2017. //FicTree: a Manually Annotated Treebank of Czech Fiction//. +
-In: J. Hlaváčová (Ed.): ITAT 2017 Proceedings, pp. 181–185. http://ceur-ws.org/Vol-1885/181.pdf+
  
-=== Acknowledgments ===+The FicTree treebank can be accessed in several ways: 
 +  - [[en:cnk:fictree#a_cnc_corpus_in_the_kontext_interface|A CNC corpus in the KonText interface]]: FicTree is available as a [[en:cnk:uvod|CNC corpus]] in the [[en:manualy:kontext:index|KonText]] interface. 
 +  - [[en:cnk:fictree#data_annotated_according_to_pdt_a-layer|Data annotated according to PDT a-layer]]: the data of the FicTree treebank annotated according to the [[http://ufal.mff.cuni.cz/pdt2.0/doc/manuals/cz/a-layer/html/index.html|PDT a-layer guidelines]] are available for download from the [[https://lindat.mff.cuni.cz/repository/xmlui/handle/11234/1-2517|LINDAT/CLARIN]] repository for non-commercial use. 
 +  - [[en:cnk:fictree#data_annotated_in_the_Universal_Dependencies_standard|Data annotated in the Universal Dependencies standard]]: the data of the FicTree treebank annotated according to the [[http://universaldependencies.org/|Universal Dependencies]] standard into which it was automatically converted are available through the [[http://universaldependencies.org/treebanks/cs_fictree/index.html|UD web page]] (for non-commercial use only).
  
-We wish to thank the participants in the annotation effort, including Milena Hnátková, Ivana Klímová, Alena Kropíková, Hana Skoumalová and Olga Zitová. +===== 1. A CNC corpus in the KonText interface ===== 
 + 
 +The FicTree corpus is available in the same way as other CNC corpora through the [[en:manualy:kontext:index|KonText]] interface. 
 + 
 +The corpus annotation is accessible through a wide range of attributes for each token. The morphological annotation and lemmatization are available using the attributes [[seznamy:tagy|tag]] and [[en:pojmy:lemma|lemma]]; additionally, the information about the POS and nominal case (if applicable) of all tokens is accessible using the attributes **pos** and **case**. 
 + 
 +The syntactic annotation of FicTree can be accessed using several positional attributes (the same as in the SYN2015 corpus): 
 +  * afun – syntactic function according to the a-layer PDT annotation 
 +  * parent – relative position of the governing token 
 +  * eparent – relative position of the nearest governing content word 
 +  * prep – lemma of a preposition governing the token (if any) 
 +  * p_lemma, p_tag, ep_lemma, ep_tag – tag and lemma of the governing token 
 +  * p_pos, p_case, ep_pos, ep_case – POS and case of the governing token 
 +  * p_afun, ep_afun – syntactic function of the governing token 
 + 
 +===== 2. Data annotated according to PDT a-layer ===== 
 + 
 +The data of the FicTree treebank, annotated according to the PDT a-layer guidelines, are available through the  
 +[[https://lindat.mff.cuni.cz/repository/xmlui/handle/11234/1-2517|LINDAT/CLARIN]] repository in a ''vertical'' format (tab-separated values), sentence boundaries are marked with empty lines. Each word is written on a single line, followed by five attributes separated by tabulators: **lemma**, **tag**, **ID** (number indicating the position of the token in the sentence), **head** (ID of the governing token) and **afun** (syntactic function according to [[http://ufal.mff.cuni.cz/pdt2.0/doc/manuals/cz/a-layer/html/index.html|PDT]]). The texts are divided into segments of maximum 100 tokens (respecting sentence boundaries), the segments are randomly shuffled. Every such segment constitues a single file, the name of the file refers to the literary work, its prefix indicates the intended partition of the data into train, dev and test data (80% - 10% - 10%). 
 + 
 +===== 3. Data annotated in the Universal Dependencies standard ===== 
 + 
 +The morphological and syntactic annotation according to the UD guidelines was performed by converting the original PDT annotation. The conversion procedure was designed by Dan Zeman and implemented in [[https://github.com/ufal/treex|Treex]]. 
 +The data are available on the [[http://universaldependencies.org/treebanks/cs_fictree/index.html|Universal Dependencies]] webpage. They are in the [[http://universaldependencies.org/format.html|CONLL-U format]]. The original texts are divided into segments of maximum 100 tokens, the segments are shuffled and divided into a train, val and test data set. The FicTree treebank in UD standard is also accessible using the query tool  [[https://lindat.mff.cuni.cz/services/pmltq/|PML-TQ]]. 
 + 
 +===== Acknowledgments ===== 
 +We wish to thank the human annotators: Ivana Klímová, Alena Kropíková and Olga Zitová; as well as Dan Zeman for the data conversion. 
 + 
 +===== How to cite FicTree ===== 
 +<WRAP round tip 70%> 
 +Jelínek, T. – Hnátková, M. – Skoumalová, H.: //FicTree: manuálně syntakticky anotovaný korpus české beletrie//. Ústav Českého národního korpusu FF UK, Praha 2017. Dostupný z WWW: http://www.korpus.cz 
 + 
 +Jelínek, T.: FicTree: a Manually Annotated Treebank of Czech Fiction. In: J. Hlaváčová (Ed.): //ITAT 2017 Proceedings//, pp. 181–185. http://ceur-ws.org/Vol-1885/181.pdf 
 +</WRAP>
  

Trace: