Semantic hierarchy construction aims to build structures of concepts linked by hypernym-hyponym ("is-a") relations.A major challenge for this task is the automatic discovery such relations.This paper proposes a novel and effective method semantic hierarchies based on word embeddings, which can be used measure relationship between words.We identify whether candidate pair has relation using word-embedding-based projections words their hypernyms.Our result, an F-score 73.74%, outperforms...

Jiang Guo, Wanxiang Che, David Yarowsky, Haifeng Wang, Ting Liu. Proceedings of the 53rd Annual Meeting Association for Computational Linguistics and 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2015.

We propose a mixture-of-experts approach for unsupervised domain adaptation from multiple sources. The key idea is to explicitly capture the relationship between target example and different source domains. This relationship, expressed by point-to-set metric, determines how combine predictors trained on various metric learned in an fashion using meta-training. Experimental results sentiment analysis part-of-speech tagging demonstrate that our consistently outperforms baselines can robustly...

Recent work has shown success in using continuous word embeddings learned from unlabeled data as features to improve supervised NLP systems, which is regarded a simple semi-supervised learning mechanism. However, fundamental problems on effectively incorporating the embedding within framework of linear models remain. In this study, we investigate and analyze three different approaches, including new proposed distributional prototype approach, for utilizing features. The presented approaches...

Yuxuan Wang, Wanxiang Che, Jiang Guo, Yijia Liu, Ting Liu. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint (EMNLP-IJCNLP). 2019.

Distant supervised relation extraction (RE) has been an effective way of finding novel relational facts from text without labeled training data. Typically it can be formalized as a multi-instance multi-label problem.In this paper, we introduce neural approach for distant with specific focus on attention mechanisms.Unlike the feature-based logistic regression model and compositional models such CNN, our includes two major attention-based memory components, which is capable explicitly...

Cross-lingual model transfer has been a promising approach for inducing dependency parsers low-resource languages where annotated treebanks are not available. The major obstacles the two-fold: 1. Lexical features directly transferable across languages; 2. Target language-specific syntactic structures difficult to be recovered. To address these two challenges, we present novel representation learning framework multi-source parsing. Our allows parsing using full lexical straightforwardly. By...

Abstract Digital polymers with precisely ordered units acting as the coded 0- or 1-bit, are introduced a promising option for molecular data storage. However, pursuit of better performance in terms high storage capacity and useful functions never stops. Herein, we propose concept an information-coded 2D digital dendrimer. The divergent growth via thiol-maleimide Michael coupling allows precise arrangements 1-bits uniform dendrimers. A protocol calculating non-linear binary dendrimer is...

With recent advances in the computer-aided synthesis planning (CASP) powered by data science and machine learning, modern CASP programs can rapidly identify thousands of potential pathways for a given target molecule. However, lack holistic pathway evaluation mechanism makes it challenging to systematically prioritize strategic except using some simple heuristics. Herein, we introduce data-driven approach evaluate relative levels retrosynthesis dynamic tree-structured long short-term memory...

Molecular structure recognition is the task of translating a molecular image into its graph structure. Significant variation in drawing styles and conventions exhibited chemical literature poses significant challenge for automating this task. In paper, we propose MolScribe, novel image-to-graph generation model that explicitly predicts atoms bonds, along with their geometric layouts, to construct Our flexibly incorporates symbolic chemistry constraints recognize chirality expand abbreviated...

This paper investigates the problem of network embedding, which aims at learning low-dimensional vector representation nodes in networks. Most existing embedding methods rely solely on structure, i.e., linkage relationships between nodes, but ignore rich content information associated with it, is common real world networks and beneficial to describing characteristics a node. In this paper, we propose content-enhanced (CENE), capable jointly leveraging structure information. Our approach...

Semantic dependency graph has been recently proposed as an extension of tree-structured syntactic or semantic representation for natural language sentences. It particularly features the structural property multi-head, which allows nodes to have multiple heads, resulting in a directed acyclic graph(DAG) parsing problem. Yet most statistical parsers focused exclusively on shallow bi-lexical tree structures, DAG remains under-explored. In this paper, we propose neural transition-based parser,...

Most modern Information Extraction (IE) systems are implemented as sequential taggers and only model local dependencies. Non-local non-sequential context is, however, a valuable source of information to improve predictions. In this paper, we introduce GraphIE, framework that operates over graph representing broad set dependencies between textual units (i.e. words or sentences). The algorithm propagates connected nodes through convolutions, generating richer representation can be exploited...

Yujie Qian, Enrico Santus, Zhijing Jin, Jiang Guo, Regina Barzilay. Proceedings of the 2019 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). 2019.

Reaction diagram parsing is the task of extracting reaction schemes from a in chemistry literature. The diagrams can be arbitrarily complex; thus, robustly them into structured data an open challenge. In this paper, we present RxnScribe, machine learning model for varying styles. We formulate prediction with sequence generation approach, which condenses traditional pipeline end-to-end model. train RxnScribe on dataset 1378 and evaluate it cross validation, achieving 80.0% soft match F1...

Many natural language processing (NLP) tasks can be generalized into segmentation problem. In this paper, we combine semi-CRF with neural network to solve NLP tasks. Our model represents a segment both by composing the input units and embedding entire segment. We thoroughly study different composition functions embeddings. conduct extensive experiments on two typical tasks: named entity recognition (NER) Chinese word (CWS). Experimental results show that our benefits from representing...

Sentiment analysis of user-generated reviews or comments on products and services in social networks can help enterprises to analyze the feedback from customers take corresponding actions for improvement. To mitigate large-scale annotations target domain, domain adaptation (DA) provides an alternate solution by learning a transferable model other labeled source domains. Existing multi-source (MDA) methods either fail extract some discriminative features that are related sentiment, neglect...

Semantic hierarchy construction aims to build structures of concepts linked by hypernym-hyponym (“is-a”) relations. A major challenge for this task is the automatic discovery such This paper proposes a novel and effective method semantic hierarchies based on continuous vector representation words, named word embeddings, which can be used measure relationship between words. We identify whether candidate pair has relation using word-embedding-based projections words their hypernyms. Our...

Wanxiang Che, Jiang Guo, Yuxuan Wang, Bo Zheng, Huaipeng Zhao, Yang Liu, Dechuan Teng, Ting Liu. Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies. 2017.

Various treebanks have been released for dependency parsing. Despite that may belong to different languages or annotation schemes, they contain syntactic knowledge is potential benefit each other. This paper presents an universal framework exploiting these multi-typed improve parsing with deep multi-task learning. We consider two kinds of as source: the multilingual and monolingual heterogeneous treebanks. Multiple are trained jointly interacted multi-level parameter sharing. Experiments on...