EMNLP 2015 Papers

We present experiments with generative models for linearization of unordered labeled syntactic dependency trees (Belz et al., 2011; Rajkumar and White, 2014). Our linearization models are derived from generative models for dependency structure (Eisner, 1996). We present a series of generative dependency models designed to capture successively more information about ordering constraints among sister dependents. We give a dynamic programming algorithm for computing the conditional probability of word orders given tree structures under these models. The models are tested on corpora of 11 languages using test-set likelihood, and human ratings for generated forms are collected for English. Our models benefit from representing local order constraints among sisters and from backing off to less sparse distributions, including distributions not conditioned on the head.

While it has been established that transitions between discourse relations are important for coherence, such information has not so far been used to aid in language generation. We introduce an approach to discourse planning for conceptto-text generation systems which simultaneously determines the order of messages and the discourse relations between them. This approach makes it straightforward to use statistical transition models, such as n-gram models of discourse relations learned from an annotated corpus. We show that using such a model significantly improves the quality of the generated text as judged by humans.

Automatic text summarization is widely regarded as the highly difficult problem, partially because of the lack of large text summarization data set. Due to the great challenge of constructing the large scale summaries for full text, in this paper, we introduce a large corpus of Chinese short text summarization dataset constructed from the Chinese microblogging website Sina Weibo, which is released to the public {this http URL}. This corpus consists of over 2 million real Chinese short texts with short summaries given by the author of each text. We also manually tagged the relevance of 10,666 short summaries with their corresponding short texts. Based on the corpus, we introduce recurrent neural network for the summary generation and achieve promising results, which not only shows the usefulness of the proposed corpus for short text summarization research, but also provides a baseline for further research on this topic.

The most successful approaches to extractive text summarization seek to maximize bigram coverage subject to a budget constraint. In this work, we propose instead to maximize semantic volume. We embed each sentence in a semantic space and construct a summary by choosing a subset of sentences whose convex hull maximizes volume in that space. We provide a greedy algorithm based on the GramSchmidt process to efficiently perform volume maximization. Our method outperforms the state-of-the-art summarization approaches on benchmark datasets.

We propose to automatically summarize student responses to reflection prompts and introduce a novel summarization algorithm that differs from traditional methods in several ways. First, since the linguistic units of student inputs range from single words to multiple sentences, our summaries are created from extracted phrases rather than from sentences. Second, the phrase summarization algorithm ranks the phrases by the number of students who semantically mention a phrase in a summary. Experimental results show that the proposed phrase summarization approach achieves significantly better summarization performance on an engineering course corpus in terms of ROUGE scores when compared to other summarization methods, including MEAD, LexRank and MMR.

We present an approach for extractive single-document summarization. Our approach is based on a weighted graphical representation of documents obtained by topic modeling. We optimize importance, coherence and non-redundancy simultaneously using ILP. We compare ROUGE scores of our system with state-of-the-art results on scientific articles from PLOS Medicine and on DUC 2002 data. Human judges evaluate the coherence of summaries generated by our system in comparision to two baselines. Our approach obtains competitive performance.

Storyline detection from news articles aims at summarizing events described under a certain news topic and revealing how those events evolve over time. It is a difficult task because it requires first the detection of events from news articles published in different time periods and then the construction of storylines by linking events into coherent news stories. Moreover, each storyline has different hierarchical structures which are dependent across epochs. Existing approaches often ignore the dependency of hierarchical structures in storyline generation. In this paper, we propose an unsupervised Bayesian model, called dynamic storyline detection model, to extract structured representations and evolution patterns of storylines. The proposed model is evaluated on a large scale news corpus. Experimental results show that our proposed model outperforms several baseline approaches.

Predicting the success of referring expressions (RE) is vital for real-world applications such as navigation systems. Traditionally, research has focused on studying Referring Expression Generation (REG) in virtual, controlled environments. In this paper, we describe a novel study of spatial references from real scenes rather than virtual. First, we investigate how humans describe objects in open, uncontrolled scenarios and compare our findings to those reported in virtual environments. We show that REs in real-world scenarios differ significantly to those in virtual worlds. Second, we propose a novel approach to quantifying image complexity when complete annotations are not present (e.g. due to poor object recognition capabitlities), and third, we present a model for success prediction of REs for objects in real scenes. Finally, we discuss implications for Natural Language Generation (NLG) systems and future directions.

Automated text summarization is aimed at extracting essential information from original text and presenting it in a minimal, often predefined, number of words. In this paper, we introduce a new approach for unsupervised extractive summarization, based on the Minimum Description Length (MDL) principle, using the Krimp dataset compression algorithm (Vreeken et al., 2011). Our approach represents a text as a transactional dataset, with sentences as transactions, and then describes it by itemsets that stand for frequent sequences of words. The summary is then compiled from sentences that compress (and as such, best describe) the document. The problem of summarization is reduced to the maximal coverage, following the assumption that a summary that best describes the original text, should cover most of the word sequences describing the document. We solve it by a greedy algorithm and present the evaluation results.

ROUGE is a widely adopted, automatic evaluation measure for text summarization. While it has been shown to correlate well with human judgements, it is biased towards surface lexical similarities. This makes it unsuitable for the evaluation of abstractive summarization, or summaries with substantial paraphrasing. We study the effectiveness of word embeddings to overcome this disadvantage of ROUGE. Specifically, instead of measuring lexical overlaps, word embeddings are used to compute the semantic similarity of the words used in summaries instead. Our experimental results show that our proposal is able to achieve better correlations with human judgements when measured with the Spearman and Kendall rank coefficients.

This paper demonstrates the effectiveness of a Long Short-Term Memory language model in our initial efforts to generate unconstrained rap lyrics. The goal of this model is to generate lyrics that are similar in style to that of a given rapper, but not identical to existing lyrics: this is the task of ghostwriting. Unlike previous work, which defines explicit templates for lyric generation, our model defines its own rhyme scheme, line length, and verse length. Our experiments show that a Long Short-Term Memory language model produces better “ghostwritten” lyrics than a baseline model.

In concept-based summarization, sentence selection is modelled as a budgeted maximum coverage problem. As this problem is NP-hard, pruning low-weight concepts is required for the solver to find optimal solutions efficiently. This work shows that reducing the number of concepts in the model leads to lower Rouge scores, and more importantly to the presence of multiple optimal solutions. We address these issues by extending the model to provide a single optimal solution, and eliminate the need for concept pruning using an approximation algorithm that achieves comparable performance to exact inference.

This paper proposes a novel approach to generate abstractive summary for multiple documents by extracting semantic information from texts. The concept of Basic Semantic Unit (BSU) is defined to describe the semantics of an event or action. A semantic link network on BSUs is constructed to capture the semantic information of texts. Summary structure is planned with sentences generated based on the semantic link network. Experiments demonstrate that the approach is effective in generating informative, coherent and compact summary.

Speech translation is conventionally carried out by cascading an automatic speech recognition (ASR) and a statistical machine translation (SMT) system. The hypotheses chosen for translation are based on the ASR system’s acoustic and language model scores, and typically optimized for word error rate, ignoring the intended downstream use: automatic translation. In this paper, we present a coarseto-fine model that uses features from the ASR and SMT systems to optimize this coupling. We demonstrate that several standard features utilized by ASR and SMT systems can be used in such a model at the speech-translation interface, and we provide empirical results on the Fisher Spanish-English speech translation corpus.

This paper focuses on language modeling with adequate robustness to support different domain tasks. To this end, we propose a hierarchical latent word language model (h-LWLM). The proposed model can be regarded as a generalized form of the standard LWLMs. The key advance is introducing a multiple latent variable space with hierarchical structure. The structure can flexibly take account of linguistic phenomena not present in the training data. This paper details the definition as well as a training method based on layer-wise inference and a practical usage in natural language processing tasks with an approximation technique. Experiments on speech recognition show the effectiveness of hLWLM in out-of domain tasks.

In this paper, a turn-taking phenomenon taxonomy is introduced, organised according to the level of information conveyed. It is aimed to provide a better grasp of the behaviours used by humans while talking to each other, so that they can be methodically replicated in spoken dialogue systems. Five interesting phenomena have been implemented in a simulated environment: the system barge-in with three variants (resulting from either an unclear, an incoherent or a sufficient user message), the feedback and the user barge-in. The experiments reported in the paper illustrate that how such phenomena are implemented is a delicate choice as their impact on the system’s performance is variable.

Recently, neural network based dependency parsing has attracted much interest, which can effectively alleviate the problems of data sparsity and feature engineering by using the dense features. However, it is still a challenge problem to sufficiently model the complicated syntactic and semantic compositions of the dense features in neural network based methods. In this paper, we propose two heterogeneous gated recursive neural networks: tree structured gated recursive neural network (Tree-GRNN) and directed acyclic graph structured gated recursive neural network (DAG-GRNN). Then we integrate them to automatically learn the compositions of the dense features for transition-based dependency parsing. Specifically, Tree-GRNN models the feature combinations for the trees in stack, which already have partial dependency structures. DAG-GRNN models the feature combinations of the nodes whose dependency relations have not been built yet. Experiment results on two prevalent benchmark datasets (PTB3 and CTB5) show the effectiveness of our proposed model.

We describe an approach to create a diverse set of predictions with spectral learning of latent-variable PCFGs (L-PCFGs). Our approach works by creating multiple spectral models where noise is added to the underlying features in the training set before the estimation of each model. We describe three ways to decode with multiple models. In addition, we describe a simple variant of the spectral algorithm for L-PCFGs that is fast and leads to compact models. Our experiments for natural language parsing, for English and German, show that we get a significant improvement over baselines comparable to state of the art. For English, we achieve the F1 score of 90.18, and for German we achieve theF1 score of 83.38.

Accurate multilingual transfer parsing typically relies on careful feature engineering. In this paper, we propose a hierarchical tensor-based approach for this task. This approach induces a compact feature representation by combining atomic features. However, unlike traditional tensor models, it enables us to incorporate prior knowledge about desired feature interactions, eliminating invalid feature combinations. To this end, we use a hierarchical structure that uses intermediate embeddings to capture desired feature combinations. Algebraically, this hierarchical tensor is equivalent to the sum of traditional tensors with shared components, and thus can be effectively trained with standard online algorithms. In both unsupervised and semi-supervised transfer scenarios, our hierarchical tensor consistently improves UAS and LAS over state-of-theart multilingual transfer parsers and the traditional tensor model across 10 different languages. 1