Dependency Tree Research Articles

A Generative Approach for Scheduling Multi-Robot Cooperative Three-Dimensional Printing

Abstract Cooperative 3D printing (C3DP) is a novel approach to additive manufacturing, where multiple printhead-carrying mobile robots work cooperatively to print the desired part. The core of C3DP is the chunk-based printing strategy in which the desired part is first split into smaller chunks and then the chunks are assigned to individual robots to print and bond. These robots will work simultaneously in a scheduled sequence to print the entire part. Although promising, C3DP lacks a generative approach that enables automatic chunking and scheduling. In this study, we aim to develop a generative approach that can automatically generate different print schedules for a chunked object by exploring a larger solution space that is often beyond the capability of human cognition. The generative approach contains (1) a random generator of diverse print schedules based on an adjacency matrix that represents a directed dependency tree structure of chunks; (2) a set of geometric constraints against which the randomly generated schedules will be checked for validation, and (3) a printing time evaluator for comparing the performance of all valid schedules. We demonstrate the efficacy of the generative approach using two case studies: a large simple rectangular bar and a miniature folding sport utility vehicle (SUV) with more complicated geometry. This study demonstrates that the generative approach can generate a large number of different print schedules for collision-free C3DP, which cannot be explored solely using human heuristics. This generative approach lays the foundation for building the optimization approach of C3DP scheduling.

Dependency-Based Relative Positional Encoding for Transformer NMT

This paper proposes a new Transformer neural machine translation model that incorporates syntactic distances between two source words into the relative position representations of the self-attention mechanism. In particular, the proposed model encodes pair-wise relative depths on a source dependency tree, which are differences between the depths of the two source words, in the encoder’s self-attention. The experiments show that our proposed model achieves 0.5 point gain in BLEU on the Asian Scientific Paper Excerpt Corpus Japanese-to-English translation task.

Automatic Detection of Important Tokens on Dependency Trees for Relation Classification

Automatic Detection of Important Tokens on Dependency Trees for Relation Classification

NLP Questions Answering Using DBpedia and YAGO

In this paper, we present results of employing DBpedia and YAGO as lexical databases for answering questions formulated in the natural language. The proposed solution has been evaluated for answering class 1 and class 2 questions (out of 5 classes defined by Moldovan for TREC conference). Our method uses dependency trees generated from the user query. The trees are browsed for paths leading from the root of the tree to the question subject. We call those paths fibers and they represent the user intention. The question analysis consists of three stages: query analysis, query breakdown and information retrieval. The aim of those stages is the detection of the entities of interest and its attributes, indicating the users’ domain of interest. The user query is then converted into a SPARQL query and sent to the DBpedia and YAGO databases. The baseline and the extended methods are presented and the quality of the solution is evaluated and discussed.

A mobile sink path planning for wireless sensor networks based on priority‐ordered dependent nonparametric trees

SummaryThe professional design of the routing protocols with mobile sink(s) in wireless sensor networks (WSNs) is important for many purposes such as maximizing energy efficiency, increasing network life, and evenly distributing load balance across the network. Moreover, mobile sinks ought to first collect data from nodes which have very important and dense data so that packet collision and loss can be prevented at an advanced level. For these purposes, the present paper proposes a new mobile path planning protocol by introducing priority‐ordered dependent nonparametric trees (PoDNTs) for WSNs. Unlike traditional clustered or swarm intelligence topology‐based routing methods, a topology which has hierarchical and dependent infinite tree structure provides a robust link connection between nodes, making it easier to reselect ancestor nodes (ANs). The proposed priority‐ordered infinite trees are sampled in the specific time frames by introducing new equations and hierarchically associated with their child nodes starting from the root node. Hence, the nodes with the highest priority and energy that belong to the constructed tree family are selected as ANs with an opportunistic approach. A mobile sink simply visits these ANs to acquire data from all nodes in the network and return to where it started. As a result, the route traveled is assigned as the mobile path for the current round. We have performed comprehensive performance analysis to illustrate the effectiveness of the present study using NS‐2 simulation environment. The present routing protocol has achieved better results than the other algorithms over various performance metrics.

Indonesian Short Essay Scoring Using Transfer Learning Dependency Tree LSTM

Indonesian Short Essay Scoring Using Transfer Learning Dependency Tree LSTM

Relation Extraction with Convolutional Network over Learnable Syntax-Transport Graph

A large majority of approaches have been proposed to leverage the dependency tree in the relation classification task. Recent works have focused on pruning irrelevant information from the dependency tree. The state-of-the-art Attention Guided Graph Convolutional Networks (AGGCNs) transforms the dependency tree into a weighted-graph to distinguish the relevance of nodes and edges for relation classification. However, in their approach, the graph is fully connected, which destroys the structure information of the original dependency tree. How to effectively make use of relevant information while ignoring irrelevant information from the dependency trees remains a challenge in the relation classification task. In this work, we learn to transform the dependency tree into a weighted graph by considering the syntax dependencies of the connected nodes and persisting the structure of the original dependency tree. We refer to this graph as a syntax-transport graph. We further propose a learnable syntax-transport attention graph convolutional network (LST-AGCN) which operates on the syntax-transport graph directly to distill the final representation which is sufficient for classification. Experiments on Semeval-2010 Task 8 and Tacred show our approach outperforms previous methods.

Multi-View Consistency for Relation Extraction via Mutual Information and Structure Prediction

Relation Extraction (RE) is one of the fundamental tasks in Information Extraction. The goal of this task is to find the semantic relations between entity mentions in text. It has been shown in many previous work that the structure of the sentences (i.e., dependency trees) can provide important information/features for the RE models. However, the common limitation of the previous work on RE is the reliance on some external parsers to obtain the syntactic trees for the sentence structures. On the one hand, it is not guaranteed that the independent external parsers can offer the optimal sentence structures for RE and the customized structures for RE might help to further improve the performance. On the other hand, the quality of the external parsers might suffer when applied to different domains, thus also affecting the performance of the RE models on such domains. In order to overcome this issue, we introduce a novel method for RE that simultaneously induces the structures and predicts the relations for the input sentences, thus avoiding the external parsers and potentially leading to better sentence structures for RE. Our general strategy to learn the RE-specific structures is to apply two different methods to infer the structures for the input sentences (i.e., two views). We then introduce several mechanisms to encourage the structure and semantic consistencies between these two views so the effective structure and semantic representations for RE can emerge. We perform extensive experiments on the ACE 2005 and SemEval 2010 datasets to demonstrate the advantages of the proposed method, leading to the state-of-the-art performance on such datasets.

Top-Down RST Parsing Utilizing Granularity Levels in Documents

Some downstream NLP tasks exploit discourse dependency trees converted from RST trees. To obtain better discourse dependency trees, we need to improve the accuracy of RST trees at the upper parts of the structures. Thus, we propose a novel neural top-down RST parsing method. Then, we exploit three levels of granularity in a document, paragraphs, sentences and Elementary Discourse Units (EDUs), to parse a document accurately and efficiently. The parsing is done in a top-down manner for each granularity level, by recursively splitting a larger text span into two smaller ones while predicting nuclearity and relation labels for the divided spans. The results on the RST-DT corpus show that our method achieved the state-of-the-art results, 87.0 unlabeled span score, 74.6 nuclearity labeled span score, and the comparable result with the state-of-the-art, 60.0 relation labeled span score. Furthermore, discourse dependency trees converted from our RST trees also achieved the state-of-the-art results, 64.9 unlabeled attachment score and 48.5 labeled attachment score.

Discontinuous Constituent Parsing with Pointer Networks

One of the most complex syntactic representations used in computational linguistics and NLP are discontinuous constituent trees, crucial for representing all grammatical phenomena of languages such as German. Recent advances in dependency parsing have shown that Pointer Networks excel in efficiently parsing syntactic relations between words in a sentence. This kind of sequence-to-sequence models achieve outstanding accuracies in building non-projective dependency trees, but its potential has not been proved yet on a more difficult task. We propose a novel neural network architecture that, by means of Pointer Networks, is able to generate the most accurate discontinuous constituent representations to date, even without the need of Part-of-Speech tagging information. To do so, we internally model discontinuous constituent structures as augmented non-projective dependency structures. The proposed approach achieves state-of-the-art results on the two widely-used NEGRA and TIGER benchmarks, outperforming previous work by a wide margin.

A BOW-Based Sentence Embedding Method for Chinese Event Identification

Event extraction is an important branch of information extraction. It involves two challenging issues: event identification and argument identification. Most approaches of event identification are trigger-based, which suffer a lot from sample imbalance, word ambiguity, low scalability, etc. To solve these problems, we attempt to explore sentence-based event identification, whose main issue is sentence representation. With the success of word embedding, it has attracted much attention to generate semantic embeddings of sentences. In this paper, we propose a simple BOW-based sentence embedding method to represent event sentences for Chinese event identification. We compute embeddings on the dependency tree and map different relations to simple arithmetic operations. We evaluate our method with a dataset of Chinese event identification and compare the result with other BOW-based methods. The results show that our approach significantly outperforms other BOW-based methods.

Targeted Sentiment Classification Based on Attentional Encoding and Graph Convolutional Networks

Targeted sentiment classification aims to predict the emotional trend of a specific goal. Currently, most methods (e.g., recurrent neural networks and convolutional neural networks combined with an attention mechanism) are not able to fully capture the semantic information of the context and they also lack a mechanism to explain the relevant syntactical constraints and long-range word dependencies. Therefore, syntactically irrelevant context words may mistakenly be recognized as clues to predict the target sentiment. To tackle these problems, this paper considers that the semantic information, syntactic information, and their interaction information are very crucial to targeted sentiment analysis, and propose an attentional-encoding-based graph convolutional network (AEGCN) model. Our proposed model is mainly composed of multi-head attention and an improved graph convolutional network built over the dependency tree of a sentence. Pre-trained BERT is applied to this task, and new state-of-art performance is achieved. Experiments on five datasets show the effectiveness of the model proposed in this paper compared with a series of the latest models.

A Heuristic Scaling Strategy for Multi-Robot Cooperative Three-Dimensional Printing

Abstract While three-dimensional (3D) printing has been making significant strides over the past decades, it still trails behind mainstream manufacturing due to its lack of scalability in both print size and print speed. Cooperative 3D printing (C3DP) is an emerging technology that holds the promise to mitigate both of these issues by having a swarm of printhead-carrying mobile robots working together to finish a single print job cooperatively. In our previous work, we have developed a chunk-based printing strategy to enable the cooperative 3D printing with two fused deposition modeling (FDM) mobile 3D printers, which allows each of them to print one chunk at a time without interfering with the other and the printed part. In this paper, we present a novel method in discretizing the continuous 3D printing process, where the desired part is discretized into chunks, resulting in multi-stage 3D printing process. In addition, the key contribution of this study is the first working scaling strategy for cooperative 3D printing based on simple heuristics, called scalable parallel arrays of robots for 3DP (SPAR3), which enables many mobile 3D printers to work together to reduce the total printing time for large prints. In order to evaluate the performance of the printing strategy, a framework is developed based on directed dependency tree (DDT), which provides a mathematical and graphical description of dependency relationships and sequence of printing tasks. The graph-based framework can be used to estimate the total print time for a given print strategy. Along with the time evaluation metric, the developed framework provides us with a mathematical representation of geometric constraints that are temporospatially dynamic and need to be satisfied in order to achieve collision-free printing for any C3DP strategy. The DDT-based evaluation framework is then used to evaluate the proposed SPAR3 strategy. The results validate the SPAR3 as a collision-free strategy that can significantly shorten the printing time (about 11 times faster with 16 robots for the demonstrated examples) in comparison with the traditional 3D printing with single printhead.

Graph Convolution Over Multiple Latent Context-Aware Graph Structures for Event Detection

Event detection is a particularly challenging problem in information extraction. The current neural network models have proved that dependency tree can better capture the correlation between candidate trigger words and related context in the sentence. However, syntactic information conveyed by the original dependency tree is insufficient for detecting trigger since the dependency tree obtained from natural language processing toolkits ignores semantic context information. Existing approaches employ a static graph structure based on original dependency tree which is incompetent in terms of distinguishing interrelations among trigger words and contextual words. So how to effectively make use of relevant information while ignoring irrelevant information from the dependency trees remains a challenging research question. To address this problem, we investigate a graph convolutional network over multiple latent context-aware graph structures to perform event detection. We exploit a multi-head attention mechanism on BERT representation and original adjacency matrix to generate multiple latent context-aware graph structures (a “dynamic cutting” strategy), which can automatically learn how to select the useful dependency information. Furthermore, we investigate graph convolutional networks with residual connections to combine the local and non-local contextual information. Experimental results on ACE2005 dataset show that our model achieves competitive performances compared with the methods based on dependency tree for event detection.

AGAT-Syntax as component of computer-based Ukrainian grammar

Abstract: The article describes functional features of the syntactic module of computer-based Ukrainian grammar AGAT. This is a linguistic type of computer-aided syntactic analysis, which provides full information about syntactic units and categories, in particular, predicativity, coordinate and subordinate clauses, the categories of subject and predicate etc. The developed linguistic software provides syntactic analysis of a whole sentence in the form of a dependency tree and indicates the types of syntactic relations and links. The AGAT-syntax task is to identify all varieties of compatibility – predicative, subordinate, and coordinate – of each word in the text. The grammatical characteristics of the phrase directly depend on which part of the language its keyword belongs to. The lexical and grammatical nature of the word determines its compatibility to the other words. Accordingly, phrases can be divided into substantive, adjective, pronouns, numeral, verbal and adverbial. Computer sub-grammars of valencies of the said parts of the language are built by us on a single principle: a lexema is indicated, preposition that participates in government and a case of a substantive word form in the shape of a two-letter code. In theory, according to their composition words combinations (phrases) are divided into simple, complex and combined. Dependency tree is built from two elements – nodes and connections. Nodes are wordforms and connections are relationships between the main element (“master”) and dependent element (“slave”). It enables to describe a configuration, a form, external parameters of a sentence but this is not sufficient to describe a sentence structure. Thus, the syntactic analysis has two levels: the first one attributes to each binary pair a type of syntactic relationships on the level of morphological way of expression of a “master”; the second level attributes to the connection a type of syntactic relationships, which include: subjective, objective, attributive, adverbial, completive and appositive modifying.. In such a way, the cycle of automated syntactic analysis of Ukrainian texts is completed by determining the syntactic word-combination, identifying a type of syntactic link and a type of relationship. It provides full range of characteristics that can be used for systemic study of semantic and syntactic problems. Keywords: automated syntactic analysis, dependency tree, syntactic relations, syntactic links.

Syntactically Meaningful and Transferable Recursive Neural Networks for Aspect and Opinion Extraction

In fine-grained opinion mining, extracting aspect terms (a.k.a. opinion targets) and opinion terms (a.k.a. opinion expressions) from user-generated texts is the most fundamental task in order to generate structured opinion summarization. Existing studies have shown that the syntactic relations between aspect and opinion words play an important role for aspect and opinion terms extraction. However, most of the works either relied on predefined rules or separated relation mining with feature learning. Moreover, these works only focused on single-domain extraction, which failed to adapt well to other domains of interest where only unlabeled data are available. In real-world scenarios, annotated resources are extremely scarce for many domains, motivating knowledge transfer strategies from labeled source domain(s) to any unlabeled target domain. We observe that syntactic relations among target words to be extracted are not only crucial for single-domain extraction, but also serve as invariant “pivot” information to bridge the gap between different domains. In this article, we explore the constructions of recursive neural networks based on the dependency tree of each sentence for associating syntactic structure with feature learning. Furthermore, we construct transferable recursive neural networks to automatically learn the domain-invariant fine-grained interactions among aspect words and opinion words. The transferability is built on an auxiliary task and a conditional domain adversarial network to reduce domain distribution difference in the hidden spaces effectively in word level through syntactic relations. Specifically, the auxiliary task builds structural correspondences across domains by predicting the dependency relation for each path of the dependency tree in the recursive neural network. The conditional domain adversarial network helps to learn domain-invariant hidden representation for each word conditioned on the syntactic structure. In the end, we integrate the recursive neural network with a sequence labeling classifier on top that models contextual influence in the final predictions. Extensive experiments and analysis are conducted to demonstrate the effectiveness of the proposed model and each component on three benchmark data sets.

Конструкції з сурядністю / однорідністю в поетичних текстах

The work considers peculiarities in conducting of automated syntactic analysis (ASA) on basis of poetical texts and as result studying the features of linear structure of Ukrainian texts, in particular, texts with constructions with parataxis and homogeneous parts. The process of text analysis includes two stages: the first stage starts with automated syntactic analysis of hierarchical structure of text, which finally builds a dependency tree. On the second stage the graph received as result of automated analysis specifies information on type or model of a sentence, in particular, with available constructions with parataxis homogeneous parts. The purpose of the work from linguistic standpoint is the analysis of parataxis/homogeneous parts in poetical texts and their relations to the content. The study of texts (1600 sentences) showed that constructions with parataxis can build three groups of relations with: homogenous parts, non-homogenous and a combination of both parts. Analysis of constructions with parataxis with available dominating syntactical level demonstrated specific features of poetical structure of texts of four analyzed authors Vinhranovskyi, Drach, Moisiienko and Stus. Project follow-up is viewed in the study of correlation between syntactical and semantic analysis of text structures, since the form of poetical text specifies of how the poetical language is shaped in sense of qualitative features. Study of construction with parataxis within both a single author’s style and an array of other authors’ styles presents interest and future prospects from the standpoint of carrying qualitative semantic analysis of parataxis with available homogenous items. It is possible since all variants of lexical meaning in the Corpora (their number reaches over 120 K) are automatically marked with special codes of taxonomic classification and with code showing relation to some or other taxon. This technology makes possible to conduct a comprehensive analysis of constructions with parataxis, to create lists and sort out relations of such type of constructions within the range of a dependency tree to detect and identify common and miscellaneous features in the structure of parataxis as a parameter of an individual writing style of a single author.

Симметрика синтаксических фигур в художественной прозе: на материале русского рассказа XX века

Статья посвящена исследованию линиаризованных соподчинительных структур, которые в теории зависимостей называются шириной куста. Исследование имеет квантитативный характер. В нем используются ранговые распределения фигур, в которых фиксируется не только число зависимых членов, но и варианты их расположения справа (в постпозиции) или слева (в препозиции) относительно вершины дерева предложения. Установлен факт сходства ранговых структур у разных авторов при существенном различии частотных характеристик. В работе рассматриваются симметрийные свойства кустовых фигур, их тяготения к зеркальной и золотой симметрии. Установлено примерное равновесие левой и правой ширины в русском рассказе первой трети XX века. Протестирован на предмет адекватности русскому языку и скорректирован закон нарастающих членов предложения немецкого филолога Отто Бехагеля.

The Interplay Between Loss Functions and Structural Constraints in Dependency Parsing

Dependency parsing can be cast as a combinatorial optimization problem with the objective to find the highest-scoring graph, where edge scores are learnt from data. Several of the decoding algorithms that have been applied to this task employ structural restrictions on candidate solutions, such as the restriction to projective dependency trees in syntactic parsing, or the restriction to noncrossing graphs in semantic parsing. In this paper we study the interplay between structural restrictions and a common loss function in neural dependency parsing, the structural hingeloss. We show how structural constraints can make networks trained under this loss function diverge and propose a modified loss function that solves this problem. Our experimental evaluation shows that the modified loss function can yield improved parsing accuracy, compared to the unmodified baseline.

A Generative Model for Punctuation in Dependency Trees

Treebanks traditionally treat punctuation marks as ordinary words, but linguists have suggested that a tree’s “true” punctuation marks are not observed (Nunberg, 1990). These latent “underlying” marks serve to delimit or separate constituents in the syntax tree. When the tree’s yield is rendered as a written sentence, a string rewriting mechanism transduces the underlying marks into “surface” marks, which are part of the observed (surface) string but should not be regarded as part of the tree. We formalize this idea in a generative model of punctuation that admits efficient dynamic programming. We train it without observing the underlying marks, by locally maximizing the incomplete data likelihood (similarly to the EM algorithm). When we use the trained model to reconstruct the tree’s underlying punctuation, the results appear plausible across 5 languages, and in particular are consistent with Nunberg’s analysis of English. We show that our generative model can be used to beat baselines on punctuation restoration. Also, our reconstruction of a sentence’s underlying punctuation lets us appropriately render the surface punctuation (via our trained underlying-to-surface mechanism) when we syntactically transform the sentence.