| { | |
| "paper_id": "Y06-1017", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T13:34:11.749641Z" | |
| }, | |
| "title": "A Chinese Dependency Syntax for Treebanking", | |
| "authors": [ | |
| { | |
| "first": "Haitao", | |
| "middle": [], | |
| "last": "Liu", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "Communication University of China", | |
| "location": { | |
| "postCode": "CN-100024", | |
| "settlement": "Beijing", | |
| "country": "China" | |
| } | |
| }, | |
| "email": "byliuhaitao@cuc.edu.cn" | |
| }, | |
| { | |
| "first": "Wei", | |
| "middle": [], | |
| "last": "Huang", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "Communication University of China", | |
| "location": { | |
| "postCode": "CN-100024", | |
| "settlement": "Beijing", | |
| "country": "China" | |
| } | |
| }, | |
| "email": "" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "This paper presents a Chinese dependency syntax for treebanking. The syntax contains 13 word classes and 34 dependency types. A format of treebank based on the syntax is also proposed for the applications of computational and general linguistic research. Some experiments show that the treebank based on the proposed dependency syntax can be used for training and evaluating the dependency parser and for quantitative analysis of Chinese syntax.", | |
| "pdf_parse": { | |
| "paper_id": "Y06-1017", | |
| "_pdf_hash": "", | |
| "abstract": [ | |
| { | |
| "text": "This paper presents a Chinese dependency syntax for treebanking. The syntax contains 13 word classes and 34 dependency types. A format of treebank based on the syntax is also proposed for the applications of computational and general linguistic research. Some experiments show that the treebank based on the proposed dependency syntax can be used for training and evaluating the dependency parser and for quantitative analysis of Chinese syntax.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Abstract", | |
| "sec_num": null | |
| } | |
| ], | |
| "body_text": [ | |
| { | |
| "text": "Treebanks are often used as a tool and resource for training and evaluating a syntactic parser in computational linguistics [1] . However, treebanks are not only useful to computational linguists, they are also an important tool for linguists from other branches of linguistics. In this way, if we are planning to build a treebank, perhaps we have to consider its applications in these two fields.", | |
| "cite_spans": [ | |
| { | |
| "start": 124, | |
| "end": 127, | |
| "text": "[1]", | |
| "ref_id": "BIBREF0" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "Although the Penn treebank based on phrase structure is still the standard of computational applications of treebank, many new projects, particularly in Europe, like to use dependency structure as the annotation schemes [5] .", | |
| "cite_spans": [ | |
| { | |
| "start": 220, | |
| "end": 223, | |
| "text": "[5]", | |
| "ref_id": "BIBREF4" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "In this paper, we present a Chinese dependency syntax for treebanking. A format of dependency treebank is proposed for training and evaluate a dependency parser and for quantitative analysis of Chinese.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "In section 2, we discuss some kernels of dependency grammar and how to build a dependency grammar for a language. Section 3 presents the proposed Chinese dependency syntax and how to process coordinating structures in this syntax. In section 4, a format of dependency treebank is given and some questions on the treebanking are discussed.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "Sentence analysis based on the dependency relations has a longer history than the method based on phrase structure. The ideas of dependency analysis are found more or less in traditional grammar of many languages. In other words, the school grammars in many countries are similar with the syntactic principles based on dependency relations. This is an important argument, if we consider that building of treebank is a labor-intensive task.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Dependency relation and dependency syntax", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "What's dependency relation, this is not a simple question, because \"It is perhaps surprising that one question which, as far as I know, has never been seriously addressed in the Dependency Grammar (DG) literature is what the dependency relation might really amount to\" [6] . Today Penn treebank is already seen as a gold standard of training and evaluating parser. However, if we want NLP to solve many more language problems for our world, only the Penn treebank is not enough, we have to build more treebanks including more languages and genres, but it is a very difficult task, because \"it is unrealistic and almost impossible to train experts in a specialized scientific domain so as to do Penn-style phrase structure annotation to the texts of their expertise.\" [15] If this is true, we have to search a more easily used scheme to annotate the treebank. Perhaps this also explains why there are more and more treebanks based on dependency relations.", | |
| "cite_spans": [ | |
| { | |
| "start": 197, | |
| "end": 201, | |
| "text": "(DG)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 269, | |
| "end": 272, | |
| "text": "[6]", | |
| "ref_id": "BIBREF5" | |
| }, | |
| { | |
| "start": 767, | |
| "end": 771, | |
| "text": "[15]", | |
| "ref_id": "BIBREF14" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Dependency relation and dependency syntax", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "It is worthwhile to notice that, although we have some native dependency treebanks, such as PDT (Prague Dependency Treebank, [3] ), many studies about dependency parsing are using a treebank converted from the Penn treebank. It is understandable to use such non-native dependency for easily comparing the results with other methods. Nevertheless, this approach is problematic, because the fundamental principles of sentence analysis in these two methods are not completely the same. For example, there are too few dependency types in the non-native dependency treebank, which is not acceptable for dependency syntactician.", | |
| "cite_spans": [ | |
| { | |
| "start": 125, | |
| "end": 128, | |
| "text": "[3]", | |
| "ref_id": "BIBREF2" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Dependency relation and dependency syntax", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "How to build a dependency syntax based on the concepts of dependency grammar tradition? Firstly, the linguist has to determine the word classes of the defined language. Secondly, he should define the dependency types (relations) of a language.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Dependency relation and dependency syntax", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "After having a set of word class and dependency types, we can construct dependency syntax of a language, which diagrammatically is presented as in figure 1. where wc 1 wc 2 ...wc m are word classes of the language and dep 1 dep 2 ...dep n are the dependency types of the language. Figure 1 gives not only the capacity of a word class governing other word classes through some dependency types, it also provides a capacity of a word class governed by other word classes. These two capacities consist of the general valency pattern of a word class and the pattern is the driven force of generating the dependency relations. Based on the valency pattern, for improving the expressing and explaining capacity of the valency patterns, a new kind of pattern with probabilistic elements is developed. Under the new name \"Weighted Valency Pattern\", the relation between two word classes will not only be described qualitatively, but also be defined quantitatively [8] .", | |
| "cite_spans": [ | |
| { | |
| "start": 956, | |
| "end": 959, | |
| "text": "[8]", | |
| "ref_id": "BIBREF7" | |
| } | |
| ], | |
| "ref_spans": [ | |
| { | |
| "start": 281, | |
| "end": 289, | |
| "text": "Figure 1", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Dependency relation and dependency syntax", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "If word class and dependency type are the core elements of dependency syntax, what is the number of the word classes and types in a language? If it is too few, the analysis and learning based on the syntax will be easier, but it will degrade the describing precision of the model. If it were too numerous, the recognizing of the type will be more difficult for human and machine, and turns down the precision and efficiency of annotating and parsing. For finding an appropriate value, we investigate dependency grammars of more than 10 languages. The table 1 shows the result 1 . ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Fig. 1. Diagrammatic dependency syntax", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Dependency syntax of a language contains two parts: the tagset of word classes, the tagset of dependency types. Our task is for building a dependency syntax to annotate a treebank, which is a time-consuming and labor-intensive task, we have to think how to automate the process and find more competent workers from early beginning. In other words, we should use the available resource as far as possible.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Chinese dependency syntax", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "Based on the national standard \"POS tagset for Chinese information processing\" (2003) and popularly used \"Grammar system for middle-school teaching\", we propose a set of word class with 13 types: According to the three basic elements consisting of dependency relation: governor, dependent and dependency type, the tagset of Chinese dependency type is built. The dependency tagset contains 20 complements and 14 adjuncts. The amount of Chinese complements is a little more than other languages, because Chinese has to use the functional words for completing the grammatical functions, which often are morphologically realized in other languages.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Chinese dependency syntax", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "noun (n), verb (v", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Chinese dependency syntax", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "We use a method similar to PDT [16] and DLT [9] to process the coordinating structure. In Chinese the often used conjunctions are he(\u548c), yu(\u4e0e), tong(\u540c) and punctuation(\u3001). Figure 3 shows a structure with 2 conjunctions and three conjuncts. ", | |
| "cite_spans": [ | |
| { | |
| "start": 31, | |
| "end": 35, | |
| "text": "[16]", | |
| "ref_id": "BIBREF15" | |
| }, | |
| { | |
| "start": 44, | |
| "end": 47, | |
| "text": "[9]", | |
| "ref_id": "BIBREF8" | |
| } | |
| ], | |
| "ref_spans": [ | |
| { | |
| "start": 172, | |
| "end": 180, | |
| "text": "Figure 3", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Fig. 3. A hierarchy of dependency types in Chinese", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "After having the dependency syntax, we have adopted the format in table 2 for a Chinese dependency treebank. ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Chinese dependency treebank", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "\u8fd9 r 2 \u662f v subj S1 2 \u662f v 6 \u3002 bjd s S1 3 \u4e09 m 4 \u4e2a q qc S1 4 \u4e2a q 5 \u4f8b\u5b50 n atr S1 5 \u4f8b\u5b50 n 2 \u662f v obj S1 6 \u3002 bjd", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Chinese dependency treebank", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "This format includes all three mentioned properties of dependency relation, but sometimes it is also helpful to construct a connected directed labeled graph [10] as in Figure 5 .", | |
| "cite_spans": [ | |
| { | |
| "start": 157, | |
| "end": 161, | |
| "text": "[10]", | |
| "ref_id": "BIBREF9" | |
| } | |
| ], | |
| "ref_spans": [ | |
| { | |
| "start": 168, | |
| "end": 176, | |
| "text": "Figure 5", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Chinese dependency treebank", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "It seems that the proposed format includes some redundant information only from the viewpoint of computational linguistics, but because we also hope to use the treebank for quantitative analysis of Chinese, current format is helpful to satisfy the mentioned two tasks. It is easy to convert the format into a more exchangeable XML format 3 .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Fig. 4. The dependency analysis as a graph", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "-<sentence id=\"1\" user=\"Liu Haitao\" date=\"2006-06-22\"> <word id=\"1\" form=\"\u8fd9\" postag=\"r\" head=\"2\" deprel=\"subj\" /> <word id=\"2\" form=\"\u662f\" postag=\"v\" head=\"6\" deprel=\"s\" /> <word id=\"3\" form=\"\u4e09\" postag=\"m\" head=\"4\" deprel=\"qc\" /> <word id=\"4\" form=\"\u4e2a\" postag=\"q\" head=\"5\" deprel=\"atr\" /> <word id=\"5\" form=\"\u4f8b\u5b50\" postag=\"n\" head=\"2\" deprel=\"obj\" /> <word id=\"6\" form=\"\u3002\" postag=\"bjd\" head=\"0\" deprel=\"ROOT\" /> </sentence>", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Fig. 4. The dependency analysis as a graph", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "For verifying the dependency syntax and the format, we have built a tentative dependency treebank 4 .", | |
| "cite_spans": [ | |
| { | |
| "start": 98, | |
| "end": 99, | |
| "text": "4", | |
| "ref_id": "BIBREF3" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Fig. 4. The dependency analysis as a graph", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The treebank is built on the news (xinwen lianbo) of China Central Television, a genre which is intended to be spoken but whose style is similar to the written language. We select four complete broadcasting news of this program as the annotated material. For finding the comfortable annotation means, we make two groups of annotator to annotate the texts. Before beginning the annotation, training about dependency grammar, how to recognize a dependency relation in particular, was given to all participators of this project. The first group consists of 20 undergraduates of Chinese linguistics, and the other group is only a graduate of computational linguistics with a background of Chinese linguistics. While everyone of the first group annotates the text with 500 tokens, the graduate has to finish the text with 10000 tokens. Then, two teachers of Chinese linguistics review all annotated texts. The final treebank includes 711 sentences and 20034 word tokens, so the mean sentence length is 28 words.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Fig. 4. The dependency analysis as a graph", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The annotators use MS-Excel or Access as the tool. It is feasible, because our treebank is very small, but these tools are not suitable for annotating long sentence. For making the annotation more easily, a system, with the name Depenendecy Grammar Annotator (DGA), has been developed [17] .", | |
| "cite_spans": [ | |
| { | |
| "start": 285, | |
| "end": 289, | |
| "text": "[17]", | |
| "ref_id": "BIBREF16" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Fig. 4. The dependency analysis as a graph", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Compared with the Penn Chinese Treebank [14] and the Sinica Treebank [2] , our treebank is a native dependency treebank, whose annotate scheme is mainly based on the tradition of dependency grammar without phrase structure. We are using a pure syntactic annotation scheme, which make a distinction with the Prague Dependency Treebank [3] and the PropBank [12] .", | |
| "cite_spans": [ | |
| { | |
| "start": 40, | |
| "end": 44, | |
| "text": "[14]", | |
| "ref_id": "BIBREF13" | |
| }, | |
| { | |
| "start": 69, | |
| "end": 72, | |
| "text": "[2]", | |
| "ref_id": "BIBREF1" | |
| }, | |
| { | |
| "start": 334, | |
| "end": 337, | |
| "text": "[3]", | |
| "ref_id": "BIBREF2" | |
| }, | |
| { | |
| "start": 355, | |
| "end": 359, | |
| "text": "[12]", | |
| "ref_id": "BIBREF11" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Fig. 4. The dependency analysis as a graph", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "We have also successfully made some experiments based on the treebank, for instance, the dependency parsing and quantitative analysis of Chinese A statistical Chinese dependency parser, using a general parser Maltparser released by Nivre [11] and the treebank built in this paper, has got the score 0.759 (UAS, unlabeled attachment score) and 0.712 (LAS, labeled attachment score). The analysis based on the treebank shows that Chinese has much in common with other languages, for example, the distribution of noun in a text; but it also shows some particularities, for example that dependency distance is much greater in Chinese than in English, German and Japanese [8] .", | |
| "cite_spans": [ | |
| { | |
| "start": 238, | |
| "end": 242, | |
| "text": "[11]", | |
| "ref_id": "BIBREF10" | |
| }, | |
| { | |
| "start": 667, | |
| "end": 670, | |
| "text": "[8]", | |
| "ref_id": "BIBREF7" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Fig. 4. The dependency analysis as a graph", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "We are annotating a spoken language corpus using the same syntax and scheme, that is a necessary complementation to the current treebank.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Fig. 4. The dependency analysis as a graph", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "In the paper, we propose a Chinese dependency syntax and a format of treebank based on the syntax. A tentative dependency treebank is built for verifying that the proposed syntax is suitable to annotate a Chinese text. The built treebank is not only used as the resource to train and evaluate the parser, also 4 Annotation guidelines in [7] .", | |
| "cite_spans": [ | |
| { | |
| "start": 310, | |
| "end": 311, | |
| "text": "4", | |
| "ref_id": "BIBREF3" | |
| }, | |
| { | |
| "start": 337, | |
| "end": 340, | |
| "text": "[7]", | |
| "ref_id": "BIBREF6" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Conclusions", | |
| "sec_num": "5" | |
| }, | |
| { | |
| "text": "132 works as a database for quantitative analysis of Chinese. The experiments show that the proposed dependency syntax and treebank are useful not only for computational linguistics, but it is also helpful for general syntactic studies.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Conclusions", | |
| "sec_num": "5" | |
| }, | |
| { | |
| "text": "The data are extracted from[9] and other technical report published by BSO/DLT project. The amount of dependency type is the sum of complements and adjuncts.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "[7] includes a detail explanation about subclass's tags.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "This is a Malt-XML format. http://w3.msi.vxu.se/~nivre/research/MaltXML.html", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [], | |
| "bib_entries": { | |
| "BIBREF0": { | |
| "ref_id": "b0", | |
| "title": "2003) Treebank: Building and using Parsed Corpora", | |
| "authors": [ | |
| { | |
| "first": "Anne", | |
| "middle": [], | |
| "last": "Abeill\u00e9", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": null, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Abeill\u00e9, Anne (ed. 2003) Treebank: Building and using Parsed Corpora. Dordrecht: Kluwer.", | |
| "links": null | |
| }, | |
| "BIBREF1": { | |
| "ref_id": "b1", | |
| "title": "Sinica Treebank: Design Criteria, Representational Issues and Implementation", | |
| "authors": [ | |
| { | |
| "first": "Keh-Jiann", | |
| "middle": [], | |
| "last": "Chen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2003, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "231--248", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Chen, Keh-Jiann et al. (2003) Sinica Treebank: Design Criteria, Representational Issues and Implementation. in Anne Abeille (ed. 2003). pp. 231-248.", | |
| "links": null | |
| }, | |
| "BIBREF2": { | |
| "ref_id": "b2", | |
| "title": "The Prague Dependency Treebank: A Three-Level Annotation Scenario", | |
| "authors": [ | |
| { | |
| "first": "Jan", | |
| "middle": [], | |
| "last": "Haji\u010d", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Alena", | |
| "middle": [], | |
| "last": "B\u00f6hmov\u00e1", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Eva", | |
| "middle": [], | |
| "last": "Haji\u010dov\u00e1", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2003, | |
| "venue": "Barbora Vidov\u00e1 Hladk\u00e1", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "103--127", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Haji\u010d, Jan, Alena B\u00f6hmov\u00e1, Eva Haji\u010dov\u00e1, Barbora Vidov\u00e1 Hladk\u00e1 (2003) The Prague Dependency Tree- bank: A Three-Level Annotation Scenario. in A. Abeill\u00e9 (ed. 2003), pp. 103-127.", | |
| "links": null | |
| }, | |
| "BIBREF3": { | |
| "ref_id": "b3", | |
| "title": "English Word Grammar", | |
| "authors": [ | |
| { | |
| "first": "R", | |
| "middle": [ | |
| "A" | |
| ], | |
| "last": "Hudson", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1990, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Hudson, R. A. (1990) English Word Grammar. Oxford: Blackwell.", | |
| "links": null | |
| }, | |
| "BIBREF4": { | |
| "ref_id": "b4", | |
| "title": "Dependency Treebanks: Methods, Annotation Schemes and Tools", | |
| "authors": [ | |
| { | |
| "first": "T", | |
| "middle": [], | |
| "last": "Kakkonen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2005, | |
| "venue": "Proceedings of the 15th Nordic Conference of Computational Linguistics", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Kakkonen, T. (2005) Dependency Treebanks: Methods, Annotation Schemes and Tools. in Proceedings of the 15th Nordic Conference of Computational Linguistics (NODALIDA 2005). Joensuu, Finland.", | |
| "links": null | |
| }, | |
| "BIBREF5": { | |
| "ref_id": "b5", | |
| "title": "Extraction, Movement and Dependency Theory", | |
| "authors": [ | |
| { | |
| "first": "Christian", | |
| "middle": [], | |
| "last": "Kreps", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1997, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Kreps, Christian (1997) Extraction, Movement and Dependency Theory. PhD thesis, University College Lon- don.", | |
| "links": null | |
| }, | |
| "BIBREF6": { | |
| "ref_id": "b6", | |
| "title": "A Chinese syntax based on dependency relations. CUC Tech", | |
| "authors": [ | |
| { | |
| "first": "Haitao", | |
| "middle": [], | |
| "last": "Liu", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2005, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Liu, Haitao (2005) A Chinese syntax based on dependency relations. CUC Tech. Report.", | |
| "links": null | |
| }, | |
| "BIBREF7": { | |
| "ref_id": "b7", | |
| "title": "Syntactic parsing based on Dependency Relations", | |
| "authors": [ | |
| { | |
| "first": "Haitao", | |
| "middle": [], | |
| "last": "Liu", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2006, | |
| "venue": "grkg/Humankybernetik", | |
| "volume": "47", | |
| "issue": "", | |
| "pages": "124--135", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Liu, Haitao (2006) Syntactic parsing based on Dependency Relations. in grkg/Humankybernetik, 47(3): 124-135.", | |
| "links": null | |
| }, | |
| "BIBREF8": { | |
| "ref_id": "b8", | |
| "title": "Metataxis in practice: dependency syntax for multilingual machine translation", | |
| "authors": [ | |
| { | |
| "first": "D", | |
| "middle": [], | |
| "last": "Maxwell", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "K", | |
| "middle": [], | |
| "last": "Schubert", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1989, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Maxwell, D. and Schubert, K. (1989) Metataxis in practice: dependency syntax for multilingual machine translation. Dordrecht: Foris.", | |
| "links": null | |
| }, | |
| "BIBREF9": { | |
| "ref_id": "b9", | |
| "title": "Dependency syntax: theory and practice", | |
| "authors": [ | |
| { | |
| "first": "I", | |
| "middle": [ | |
| "A" | |
| ], | |
| "last": "Mel'cuk", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1988, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Mel'cuk, I. A. (1988) Dependency syntax: theory and practice. Albany: State University Press of New York.", | |
| "links": null | |
| }, | |
| "BIBREF10": { | |
| "ref_id": "b10", | |
| "title": "Inductive Dependency Parsing", | |
| "authors": [ | |
| { | |
| "first": "J", | |
| "middle": [], | |
| "last": "Nivre", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2006, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Nivre, J. (2006) Inductive Dependency Parsing. Dordrecht: Springer.", | |
| "links": null | |
| }, | |
| "BIBREF11": { | |
| "ref_id": "b11", | |
| "title": "The Proposition Bank: A Corpus Annotated with Semantic Roles", | |
| "authors": [ | |
| { | |
| "first": "Martha", | |
| "middle": [], | |
| "last": "Palmer", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Dan", | |
| "middle": [], | |
| "last": "Gildea", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Paul", | |
| "middle": [], | |
| "last": "Kingsbury", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2005, | |
| "venue": "Computational Linguistics", | |
| "volume": "31", | |
| "issue": "1", | |
| "pages": "71--106", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Palmer, Martha, Dan Gildea, and Paul Kingsbury (2005) The Proposition Bank: A Corpus Annotated with Semantic Roles, Computational Linguistics, 31(1): 71-106.", | |
| "links": null | |
| }, | |
| "BIBREF12": { | |
| "ref_id": "b12", | |
| "title": "El\u00e9ments de la syntaxe structurale", | |
| "authors": [ | |
| { | |
| "first": "L", | |
| "middle": [], | |
| "last": "Tesni\u00e8re", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1959, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Tesni\u00e8re, L. (1959) El\u00e9ments de la syntaxe structurale. Paris: Klincksieck.", | |
| "links": null | |
| }, | |
| "BIBREF13": { | |
| "ref_id": "b13", | |
| "title": "The Penn Chinese TreeBank: Phrase Structure Annotation of a Large Corpus", | |
| "authors": [ | |
| { | |
| "first": "Nianwen", | |
| "middle": [], | |
| "last": "Xue", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Fei", | |
| "middle": [], | |
| "last": "Xia", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Fu-Dong", | |
| "middle": [], | |
| "last": "Chiou", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Martha", | |
| "middle": [], | |
| "last": "Palmer", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2005, | |
| "venue": "Natural Language Engineering", | |
| "volume": "11", | |
| "issue": "2", | |
| "pages": "207--238", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Xue, Nianwen, Fei Xia, Fu-Dong Chiou, and Martha Palmer (2005) The Penn Chinese TreeBank: Phrase Structure Annotation of a Large Corpus. Natural Language Engineering, 11(2): 207-238.", | |
| "links": null | |
| }, | |
| "BIBREF14": { | |
| "ref_id": "b14", | |
| "title": "Statistical dependency analysis with Support Vector Machines", | |
| "authors": [ | |
| { | |
| "first": "Hiroyasu", | |
| "middle": [], | |
| "last": "Yamada", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Yuji", | |
| "middle": [], | |
| "last": "Matsumoto", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2003, | |
| "venue": "Proc. 8th International Workshop on Parsing Technologies (IWPT)", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "195--206", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Yamada, Hiroyasu and Yuji Matsumoto (2003) Statistical dependency analysis with Support Vector Ma- chines. in Proc. 8th International Workshop on Parsing Technologies (IWPT), pp.195-206, April 2003.", | |
| "links": null | |
| }, | |
| "BIBREF15": { | |
| "ref_id": "b15", | |
| "title": "Valency Lexicon of Czech Verbs", | |
| "authors": [ | |
| { | |
| "first": "Zden\u011bk", | |
| "middle": [], | |
| "last": "\u017dabokrtsk\u00fd", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2005, | |
| "venue": "Faculty of Mathematics and Physics", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "\u017dabokrtsk\u00fd, Zden\u011bk (2005) Valency Lexicon of Czech Verbs. Ph.D. Thesis, Faculty of Mathematics and Phys- ics, Charles University in Prague.", | |
| "links": null | |
| }, | |
| "BIBREF16": { | |
| "ref_id": "b16", | |
| "title": "Implementation of Dependency Grammar Annotator", | |
| "authors": [ | |
| { | |
| "first": "Wei", | |
| "middle": [], | |
| "last": "Zuo", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2005, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Zuo, Wei (2005) Implementation of Dependency Grammar Annotator. Master thesis, Communication Univer- sity of China.", | |
| "links": null | |
| } | |
| }, | |
| "ref_entries": { | |
| "FIGREF0": { | |
| "text": "), adjective (a), adverb (d), pronoun (r), preposition (p), numeral (m), classifier (q), conjunction(c), interjection (e), particle (u), onomatopoeia (o) and punctuation (bd). Compared with the national standard of POS tags, we remove some tags, which do not work on the level of syntax. Comparatively with traditional school grammar, we gives some functional (particle) words an important position in the syntax, because they often play a crucial role during determining the dependency relation between two words.Figure 2shows a hierarchy of POS tagset in Chinese 2 .", | |
| "num": null, | |
| "uris": null, | |
| "type_str": "figure" | |
| }, | |
| "FIGREF1": { | |
| "text": "A hierarchy of word classes in Chinese", | |
| "num": null, | |
| "uris": null, | |
| "type_str": "figure" | |
| }, | |
| "FIGREF2": { | |
| "text": "Coordinating structureIn a coordinating structure, the first conjunction works as the head of the whole structure to connect with the head and dependents of the other structures. The prefix C-is introduced for replicating the de-pendency type which governs the whole structure. For example, in the sentence \"\u7f8e\u4e3d\u7684\u957f\u6c5f\u3001\u9ec4\u6cb3\u548c \u9ed1\u9f99\u6c5f\u90fd\u662f\u4e2d\u56fd\u7684\u6cb3\u6d41\"(The beautiful Yangzi, Huanghe and Heilongjiang are the rivers of China), the Chinese character string \"\u957f\u6c5f \u3001 \u9ec4\u6cb3 \u548c \u9ed1\u9f99\u6c5f\" forms a coordinating structure. Where Rel-SUBJ, Rel2-ATR, CC1-'\u3001', CC2-\u548c, C-Rel-C-SUBJ, C1-\u957f\u6c5f, C2-\u9ec4\u6cb3, C3-\u9ed1\u9f99\u6c5f.", | |
| "num": null, | |
| "uris": null, | |
| "type_str": "figure" | |
| }, | |
| "TABREF1": { | |
| "html": null, | |
| "type_str": "table", | |
| "content": "<table><tr><td>Language</td><td colspan=\"3\">Word class Complements Adjuncts</td></tr><tr><td>English</td><td>11</td><td>14</td><td>11</td></tr><tr><td>German</td><td>10</td><td>17</td><td>9</td></tr><tr><td>Danish</td><td>11</td><td>9</td><td>6</td></tr><tr><td>Polish</td><td>10</td><td>10</td><td>8</td></tr><tr><td>Bangla</td><td>14</td><td>10</td><td>10</td></tr><tr><td>Finnish</td><td>14</td><td>12</td><td>9</td></tr><tr><td>Hungarian</td><td>10</td><td>11</td><td>10</td></tr><tr><td>Japanese</td><td>19</td><td>13</td><td>7</td></tr><tr><td>Esperanto</td><td>12</td><td>8</td><td>10</td></tr><tr><td>French</td><td>10</td><td>11</td><td>10</td></tr><tr><td>Italian</td><td>9</td><td>10</td><td>7</td></tr><tr><td>Chinese</td><td>11</td><td>19</td><td>16</td></tr></table>", | |
| "text": "The amount of word class and dependency type of some languages The table shows that there are often about 10-20 word classes and 15-35 dependency types in the dependency syntax of a language.", | |
| "num": null | |
| }, | |
| "TABREF2": { | |
| "html": null, | |
| "type_str": "table", | |
| "content": "<table><tr><td>Type</td><td>Label</td><td>Type</td><td>Label</td></tr><tr><td>Main governor</td><td>S</td><td>Sentential object</td><td>SentObj</td></tr><tr><td>Subject</td><td>SUBJ</td><td>Auxiliary verb</td><td>ObjA</td></tr><tr><td>Object</td><td>OBJ</td><td>Coordinating mark</td><td>C-</td></tr><tr><td>Indirect Object</td><td>OBJ2</td><td>Adverbial</td><td>AVDA</td></tr><tr><td>Subobject</td><td>SUBOBJ</td><td>Verb adjunct</td><td>VA</td></tr><tr><td>Subject Complement</td><td>SOC</td><td>Attributer</td><td>ATR</td></tr><tr><td>Prepositional Object</td><td>POBJ</td><td>Topic</td><td>TOP</td></tr><tr><td>Postpositional Complement</td><td>FC</td><td>Coordinating adjunct</td><td>COOR</td></tr></table>", | |
| "text": "Chinese dependency types", | |
| "num": null | |
| }, | |
| "TABREF3": { | |
| "html": null, | |
| "type_str": "table", | |
| "content": "<table><tr><td>Order number of Sentence</td><td>Word number Character POS Order</td><td>Order number</td><td>Governor Character POS</td><td>Dependency type</td></tr><tr><td>S1</td><td>1</td><td/><td/><td/></tr></table>", | |
| "text": "Annotation of a sample sentence in treebank", | |
| "num": null | |
| } | |
| } | |
| } | |
| } |