Dependency Grammar Research Articles

Between Syntactic Carpentry and Dependency Grammar Similarities and Differences

Between Syntactic Carpentry and Dependency Grammar Similarities and Differences

A hybrid approach to extract conceptual diagram from software requirements

Employing rules for the automatic extraction of conceptual diagrams from software requirements has been in practice for some time. However, considering only rules for extraction makes the system complex to handle. Moreover, the rules are predominantly based on the syntactic structure such as Part of Speech tags along with Dependency Grammar of sentences and rarely on semantics. In this paper, we propose to use a probabilistic approach in configuration with the rule-based technique and the Word embeddings to preserve the semantics of the sentence. Hence, reduces the complexity of the extraction procedure. Further, we advocate the use of a divide-and-conquer policy of extraction instead of extracting classes for one entire use case description. We extract the class diagram from small use cases and then merge it to obtain the class diagram. As generated class diagram corresponding to small use cases can be utilized in another similar software design, thus, it increases the scalability and decreases the extraction time. The proposed hybrid approach integrates the knowledge from the experiences. Thus, the proposed approach achieved 90% as F1-score whereas the F1-Score for the existing methods ranged between 79-88%. The proposed hybrid approach also shows a 19.44% reduction in terms of the number of iterations performed to carry out extraction procedures for individual use cases. Hence, reduces the extraction procedure complexity.

Investigating the Hierarchical Relationship Between Clause and Phrase Using the Menzerath-Altmann Law: Evidence from Academic Research Articles

ABSTRACT The study justifies phrases (instead of words) as the direct constituents of clauses using English academic research articles (RAs) by fitting the Menzerath-Altmann Law (MAL) to the correlations between sentence length (in clause numbers) and clause length (in phrase numbers), and between clause length and phrase length (in word numbers). The study was conducted under the framework of dependency grammar, with a self-built corpus of 104,499 tokens. The results indicate that 1) both correlations abide by the MAL, or abide when excluding the second MAL regime, regardless of part-genres, suggesting that phrases are the appropriate direct constituents of clauses. 2) The increasing MAL fitting curve is found in , which has the shortest mean dependency distance. 3) Clustering analysis on the MAL parameters in dependency structure can effectively differentiate the size-restricted from the other part-genres at the lower level, i.e., clause-phrase-word level. Furthermore, our findings have important implications: 1) The second regime may be closely related to text characteristics or linguistic unit levels. 2) The dependency distance minimization effect is possibly a counteractive force against the Menzerathian shortening effect. 3) The competing relationship between clausal and phrasal complexity in RAs is driven by the self-regulating and self-adapting language system.

Mean dependency length — a new metric for requirements quality

AbstractThis paper proposes the mean dependency length (MDL) as a metric for measuring natural language requirements quality. Dependency length is a linguistic feature based on dependency grammar, which natural language researchers have traditionally used to evaluate syntactic complexity in other contexts. In this study, aided by MATLAB‐based algorithms, the authors assessed MDL over a requirements set composed of 249 original statements, rephrased into five pattern systems. Null hypothesis and effect size testings revealed that MDL is sensitive to the application of pattern rules and to the differences among the patterns, both in an absolute approach and in comparison with other metrics. Furthermore, it was also demonstrated that MDL is aligned with users' values, especially for understandability issues, and can be measured automatically. Finally, the work concluded that MDL is a convenient metric for assessing the quality of natural language requirements.

On the unified representation of continuity and discontinuity and its neurocognitive grounding

Abstract Discontinuity in natural language is characterized by the linear disruption of a continuous string of linguistic expressions forming a constituent. While dependency relations in Dependency Grammar (DG) can capture discontinuity well, phrase-structure-based approaches such as Phrase Structure Grammar (PSG) face difficulty in accommodating discontinuity. Categorial Grammar (CG) has correspondences with PSG, although it can handle discontinuity, if equipped with wrapping operations. Given the existing literature on discontinuity in natural language, it appears that constituency relations of PSG, dependency relations of DG and functor-argument relations of CG are distinct and independent. Here, we argue for a unified representation achieved by taking into account fundamental representational principles of PSG, DG and CG. For simplicity, we show this by considering an embedded clause from Wan, spoken in Ivory Coast, as an illustrative case. The paper then attempts to explain, based on available empirical pieces of evidence, the plausible connections between the unified representation and the neurocognitive representation of continuity and discontinuity in natural language.

Coordination and referential dependencies: a dependency grammar account in terms of predicate-valent structures

Abstract The present paper explores the nexus of coordination and referential dependencies, focusing on sentences such as ‘Max and Lucie talked about him’, in which it is hardly possible for Max and him to be co-valued. Previous accounts claim that such referential dependencies are in fact possible on a collective reading. It will be demonstrated here, however, that informants consistently judge the co-valued reading in such cases to be either strongly marginal or unavailable, regardless of whether a collective or distributive reading obtains. It is claimed that the manner in which referential dependencies interact with coordinate structures can be addressed insightfully in terms of ‘predicate-valent’ structures (as opposed to ‘predicate-argument structures’). A constraint in this regard will be proposed: the ‘Conjunct Referential Dependency Constraint’ (CRDC). The CRDC is couched in a Dependency Grammar (DG) approach to syntax.

The learnability and emergence of dependency structures in an artificial language

AbstractIn a pair of artificial language experiments, we investigated the learnability and emergence of different dependency structures: branching, center-embedding, and crossed. In natural languages, branching is the most common dependency structure; center-embedding occurs but is often disfavored, and crossed dependencies are very rare. Experiment 1 addressed learnability, testing comprehension, and production on small artificial languages exemplifying each dependency type in noun phrases. As expected, branching dependency grammars were the easiest to learn, but crossed grammars were not different from center-embedding. Experiment 2 employed iterated learning to examine the emergence and stabilization of consistent grammar using the same type of stimuli as Experiment 1. The initial participant in each chain of transmission was trained on phrases generated by a random grammar, with the language produced by that participant passed to the next participant through an iterated learning process. Branching dependency grammar appeared in most chains within a few generations and remained stable once it appeared, although one chain stabilized on output consistent with a crossed grammar; no chains converged on center-embedding grammars. These findings, along with some previous results, call into question the assumption that crossed dependencies are more cognitively complex than center-embedding, while confirming the role of learnability in the typology of dependency structures.

A Framework for Malay Computational Grammar Formalism based-on Enhanced Pola Grammar

In the era of IR4.0, Natural Language Processing (NLP) is one of the major focuses because text is stored digitally to code the information. Natural language understanding requires a computational grammar for syntax and semantics of the language in question for this information to be manipulated digitally. Many languages around the world have their own computational grammars for processing syntax and semantics. However, when it comes to the Malay language, the researchers have yet to come across a substantial computational grammar that can process Malay syntax and semantics based on a computational theoretical framework that can be applied in systems such as e-commerce. Hence, we intend to propose a formalism framework based on enhanced Pola Grammar with syntactic and semantic features. The objectives of this proposed framework are to create a linguistic computational formalism for the Malay language based on theoretical linguistic; implement templates for Malay words to handle syntax and semantic features in accordance with the enhanced Pola Grammar; and create a Malay Language Parser Algorithm that can be used for digital applications. To accomplish the objectives, the proposed framework will recursively formalise the computational Malay grammar and lexicon using a combination of solid theoretical linguistic foundations such as Dependency Grammar. A Malay parsing algorithm will be developed for the proposed model until the formalised grammar is deemed reliable. The findings of this indigenous Malay parser will help to advance Malay language applications in the digital economy.

Verbs as Central Elements in Role and Reference Grammar and Dependency Grammar and Valency Theories

Purpose: The study takes into consideration the role a verb plays in the English language, how it has been described by the traditional grammar approach, the deficiencies in their definitions, and a critical look at how verbs play the central roles in two grammatical theories: Role and Reference Grammar as well as Dependency Grammar and Valency.
 Design/Methodology/Approach: The study used the review approach to analyse texts based on examples. It relies on a thorough examination of existing literature on the topic. Content analysis was adopted to analyse the study.
 Findings: From the analysis used in the study, it was established that verbs were seen in both theories to be at the heart of meaning.
 Research Limitation: The study discusses the research limitations of both theories, including issues related to cross-linguistic differences and the difficulty of accounting for idiosyncratic uses of verbs.
 Practical Implication: The practical implications of these theories are explored, such as their usefulness in language teaching and the development of natural language processing systems.
 Social Implication: The social implications of these theories are also considered, including the potential for these theories to contribute to a better understanding of language diversity and promote linguistic equality.
 Originality/Value: The study contributes to a deeper understanding of these theoretical frameworks and their potential to enhance language research, teaching, and societal understanding.

A Quantitative Approach to the Study on Length-Frequency Relationship of English Patterns

As a master of lexicology and corpus-driven linguistic studies, pattern grammar shares the features of concern on collocation and approximation to natural language, which provides an optimization idea for the extensively used phrase structure grammar and dependency grammar in natural language processing. This research adopts the methods of quantitative linguistics, taking patterns as the research focus to explore the length-frequency relationship of patterns, verify Zipf law and synergic linguistic model in respect of patterns, and analyze the usage characteristics of patterns in different registers. The results have demonstrated that Zipf's law is also applicable to patterns, and the synergic linguistic model shows a preferable goodness of fit on the description of the pattern length-frequency relations. This study provides bidirectional contributions to pattern grammar and the synergic relationship between length and frequency. In addition, it offers a refinement method for the application of patterns in different registers.

Notational Variants and Cognition: The Case of Dependency Grammar

In recent years, dependency grammars have established themselves as valuable tools in theoretical and computational linguistics. To many linguists, dependency grammars and the more standard constituency-based formalisms are notational variants. We argue that, beyond considerations of formal equivalence, cognition may also serve as a background for a genuine comparison between these different views of syntax. In this paper, we review and evaluate some of the most common arguments and evidence employed to advocate for the cognitive or neural reality of dependency grammars in linguistics, psycholinguistics, or neurolinguistics. We then raise the possibility that the abilities to represent and track, alternatively or in parallel, constituency and dependency structures co-exist in human cognition and are constitutive of syntactic competence.

An investigation of the diachronic trend of dependency distance minimization in magazines and news.

The principle of minimization of dependency distance (DD) can reduce the working memory burden of language speakers, that is, reduce the cognitive burden during the communication process. This investigation demonstrated the dependency distance (based on the dependency grammar) minimization principle from a diachronic perspective in two text types of magazines and news with inspection indicators of mean dependency distance (MDD) and normalized dependency distance (NDD). This research revealed a fluctuation tendency around a certain axis concerning diachronic dependency distance variation. This research also indicated that news text balances language complexity and communication efficiency better than magazine text.

Deep Learning Using Context Vectors to Identify Implicit Aspects

Aspects extraction is the key task in the sentiment analysis problem, which includes extraction of both explicit and implicit aspects. Identifying implicit aspects is not a new task in sentiment analysis, but it still presents many challenges. Numerous studies have addressed the issue and offered different approaches, but there are still fundamental challenges as follows: solving the context word problem; the domain-specific problem; and the implicit aspects represented by sentiment words that do not appear explicitly in the text. This paper proposes a method for identifying implicit aspects of sentiment words. To address these above challenges, the system is built on the foundation of deep learning with context vectors, dependency grammar, anaphora coreference resolution, and sentiment ontology. The dependency parser and the entity coreference resolver are adopted to filter out the sentiment–aspect pairs data generated from embedding context words. Then, a fine-tuned technique is applied, and the classifier corresponding to each sentiment word is built to identify implicit aspects. This combination has produced positive results, reaching almost 90% accuracy.

Revisiting English written VP-ellipsis and VP-substitution: a dependency-based analysis

Abstract The present study quantitatively investigates the structural features of English written VP-ellipsis (VPE) and VP-substitution (VPS) under the theoretical framework of dependency grammar. Results show that: (a) Both VPE and VPS are human language mechanisms that follow the least effort principle; (b) Dependency distance of segmental and discontinuous linguistic units under the influence of VPE and VPS conform to the right-truncated Zipf-Alekseev distribution; (c) Dependency-based indices are effective in distinguishing similar structures of VPE and VPS, especially the Do-Support in ellipsis and Do + X in substitution. The study not only explores the regularity of partial segments of human language under the mechanisms of VPE and VPS, but also offers quantitative evidence for distinguishing confusing structures in VPE and VPS, which may aid in raising the retrieving precision of similar linguistic structures in corpora.

A Development Study on the Ordering Distribution of Temporal Adverbial Clauses by Chinese EFL Learners Based on Dependency Treebank

Abstract Temporal adverbial clause is an important language structure and exhibits different features in English and Chinese, which brings about difficulties for Chinese EFL learners. Based on the theory of Dependency Grammar, the study attempts to investigate the ordering distribution of temporal adverbial clauses by Chinese EFL learners at the beginning, intermediate and advanced levels. The results show that: 1) Chinese EFL learners at different proficiencies tend to precede temporal adverbial clause to main clause. With the increase of proficiency, the postposition of temporal adverbial clauses by learners increases and is approaching to the ordering preference of target language. 2) The ordering distribution of subordinators for temporal adverbial clauses by Chinese EFL learners is consistent with native English, showing a tendency of 100% preposition, which ascribes to the high frequency and salience of subordinators in English. 3) MDD is one of the significant motivations that cause the preference of prepositional temporal adverbial clauses by Chinese EFL learners. As a kind of natural language, interlanguage has a unique cognitive mechanism which distinguishes from both native and target language. This study provides a more comprehensive theoretical reference for learners at different proficiencies to understand and learn temporal adverbial clauses, as well as data support from empirical research for language teaching.

The Menzerath-Altmann law on the clause level in English texts

Abstract Menzerath-Altmann law (MAL) describes the relationship between the size of the construct and of its constituents, where the larger the whole, the smaller its parts. Despite numerous investigations dedicated to MAL, few studies have observed the relationship syntactically, especially at the clause level. The present study investigates three units in which clauses in English can be measured, i.e., argument, phrase, and word, by fitting MAL to the relationship between the size of the clause and its constituents. Results show that 1) clause length in phrases can be well fitted by probability distributions, while the goodness-of-fit is less favorable for clause length in arguments and words. 2) MAL holds reasonably well between the size of the clause in phrases and of the phrase in words under some conditions, i.e., within a specific range of construct size and text genres. 3) To summarize, the phrase, a notion proposed by Mačutek, Ján, Radek Čech & Jiří Milička. 2017. Menzerath-Altmann law in syntactic dependency structure. In Proceedings of the fourth international conference on dependency linguistics (Depling 2017), 100–107 under the theoretical framework of dependency grammar, is the most appropriate one to be the neighboring unit of the clause among the three measurement units. These findings may shed light on the features of syntactic structures and lead to a better understanding of the human language system.

Computing semantic similarity of texts based on deep graph learning with ability to use semantic role label information

We propose a deep graph learning approach for computing semantic textual similarity (STS) by using semantic role labels generated by a Semantic Role Labeling (SRL) system. SRL system output has significant challenges in dealing with graph-neural networks because it doesn't have a graph structure. To address these challenges, we propose a novel SRL graph by using semantic role labels and dependency grammar. For processing the SRL graph, we proposed a Deep Graph Neural Network (DGNN) based graph-U-Net model that is placed on top of the transformers to use a variety of transformers to process representations obtained from them. We investigate the effect of using the proposed DGNN and SRL graph on the performance of some transformers in computing STS. For the evaluation of our approach, we use STS2017 and SICK datasets. Experimental evaluations show that using the SRL graph accompanied by applying the proposed DGNN increases the performance of the transformers used in the DGNN.

ASK-RoBERTa: A pretraining model for aspect-based sentiment classification via sentiment knowledge mining

The main objective of aspect-based sentiment classification (ABSC) is to predict sentiment polarities of different aspects from sentences or documents. Recent research integrates sentiment terms into pretraining models whose accuracy impacts the ABSC performance. This paper introduces a sentiment knowledge-adaptive pretraining model (ASK-RoBERTa). A sentiment word dictionary is first built from general and field sentiment words. We develop a series of term and sentiment mining rules based on part-of-speech tagging and sentence dependency grammar. These mining rules consider word dependencies, compounding, and conjunctions. The pretraining model optimizes the mining rules to capture the dependency between aspects and sentiment words. Experimental results on multiple public benchmark datasets demonstrate the satisfactory performance of ASK-RoBERTa.

Syntactic and typological properties of translational language: A comparative description of dependency treebank of academic abstracts

Translational language, the language of translated texts, is distinct from both the source and the target language. Although there are many studies on translational language, they present contradictory results. This paper aims to investigate the syntactic and typological properties of translational language by adopting the two main indices of dependency grammar: mean dependency distance (MDD) and dependency direction. A comparable dependency treebank, consisting of translated and non-translated English abstracts, was built and quantitatively described. The results show that (1) the MDD of translational English is significantly longer than that of non-translational English, which can be explained by the difference in MDD of the main grammatical relations (subject, object, attribute and adverbial); (2) the MDD of translational English is within the threshold of four; (3) translational English has more head-final and fewer head-initial dependencies, which can be explained by the distributional differences of the four grammatical relations; (4) translational English has the same word order preference for SV-VO-AdjN as non-translational English, but there are more AdjN and fewer SV pairs in translational English. These findings show the effect of source language interference and suggest that a quantitative method is valid for a holistic and systemic description of translational language.

Syntactic complexity in Japanese dialects

This study attempted the clustering of Japanese dialects at a syntactic level in light of dependency grammar. it is Kagoshima, a southwestern area of Japan, that bears the longest mean dependency distance (1.7851) and Akita of the north-eastern prefecture that shows the shortest mean dependency distance (1.1805). A test of Euclidean distances and clustering based on the MDD brings to light that, the north-eastern and north Kinki areas tend to bear a shorter MDD, indicating a less complex syntax in spoken dialect; the Kantoo and Chuubu, south Kinki, and Shikoku areas tend to present a medium syntactic complexity; the south areas tend to present long MDDs. In a broader sense, the more northward the region, the deeper the syntactic complexity of spoken dialect. The longest DD (3.4) is detected in Gifu prefecture. This has to do with its location, between eastern and western Japan, connected to seven prefectures: Toyama, Ishikawa, Fukui, Shiga, Aichi, Mie and Nagano.