ACL 2016 Papers

We provide a solution for elementary science test using instructional materials. We posit that there is a hidden structure that explains the correctness of an answer given the question and instructional materials and present a unified max-margin framework that learns to find these hidden structures (given a corpus of question-answer pairs and instructional materials), and uses what it learns to answer novel elementary science questions. Our evaluation shows that our framework outperforms several strong baselines.

We explore the applicability of machine translation evaluation (MTE) methods to a very different problem: answer ranking in community Question Answering. In particular, we adopt a pairwise neural network (NN) architecture, which incorporates MTE features, as well as rich syntactic and semantic embeddings, and which efficiently models complex non-linear interactions. The evaluation results show state-of-the-art performance, with sizeable contribution from both the MTE features and from the pairwise NN architecture.

We propose a novel vector representation that integrates lexical contrast into distributional vectors and strengthens the most salient features for determining degrees of word similarity. The improved vectors significantly outperform standard models and distinguish antonyms from synonyms with an average precision of 0.66-0.76 across word classes (adjectives, nouns, verbs). Moreover, we integrate the lexical contrast vectors into the objective function of a skip-gram model. The novel embedding outperforms state-of-the-art models on predicting word similarities in SimLex-999, and on distinguishing antonyms from synonyms.

Nowadays, many scholarly messages are posted on Chinese microblogs and more and more researchers tend to find scholarly information on microblogs. In order to exploit microblogging to benefit scientific research, we propose a scholarly microblog recommendation system in this study. It automatically collects and mines scholarly information from Chinese microblogs, and makes personalized recommendations to researchers. We pro-pose two different neural network models which learn the vector representations for both users and microblog texts. Then the recommendation is accomplished based on the similarity between a user’s vector and a microblog text’s vector. We also build a dataset for this task. The two embedding models are evaluated on the dataset and show good results compared to several baselines.

Using corpus data of spoken dialogue, we examine the convergence of syntactic complexity levels between interlocutors in natural conversations, as it occurs within spans of topic episodes. The findings of general convergence in the Switchboard and BNC corpora are compatible with an information-theoretic model of dialogue and with Interactive Alignment Theory.

This paper presents a novel approach to automated sentence completion based on pointwise mutual information (PMI). Feature sets are created by fusing the various types of input provided to other classes of language models, ultimately allowing multiple sources of both local and distant information to be considered. Furthermore, it is shown that additional precision gains may be achieved by incorporating feature sets of higher-order n-grams. Experimental results demonstrate that the PMI model outperforms all prior models and establishes a new state-of-the-art result on the Microsoft Research Sentence Completion Challenge

Modern coreference resolution systems require linguistic and general knowledge typically sourced from costly, manually curated resources. Despite their intuitive appeal, results have been mixed. In this work, we instead implement fine-grained surface-level features motivated by cognitive theory. Our novel fine-grained feature specialisation approach significantly improves the performance of a strong baseline, achieving state-of-the-art results of 65.29 and 61.13% on CoNLL-2012 using gold and automatic preprocessing, with system extracted mentions.

We introduce an approach to train lexicalized parsers using bilingual corpora obtained by merging harmonized treebanks of different languages, producing parsers that can analyze sentences in either of the learned languages, or even sentences that mix both. We test the approach on the Universal Dependency Treebanks, training with MaltParser and MaltOptimizer. The results show that these bilingual parsers are more than competitive, as most combinations not only preserve accuracy, but some even achieve significant improvements over the corresponding monolingual parsers. Preliminary experiments also show the approach to be promising on texts with code-switching and when more languages are added.

In this paper, we propose LexVec, a new method for generating distributed word representations that uses low-rank, weighted factorization of the Positive Point-wise Mutual Information matrix via stochastic gradient descent, employing a weighting scheme that assigns heavier penalties for errors on frequent co-occurrences while still accounting for negative co-occurrence. Evaluation on word similarity and analogy tasks shows that LexVec matches and often outperforms state-of-the-art methods on many of these tasks.

Bidirectional long short-term memory (bi-LSTM) networks have recently proven successful for various NLP sequence modeling tasks, but little is known about their reliance to input representations, target languages, data set size, and label noise. We address these issues and evaluate bi-LSTMs with word, character, and unicode byte embeddings for POS tagging. We compare bi-LSTMs to traditional POS taggers across languages and data sizes. We also present a novel bi-LSTM model, which combines the POS tagging loss function with an auxiliary loss function that accounts for rare words. The model obtains state-of-the-art performance across 22 languages, and works especially well for morphologically complex languages. Our analysis suggests that bi-LSTMs are less sensitive to training data size and label corruptions (at small noise levels) than previously assumed.

Phrase table pruning is the act of removing phrase pairs from a phrase table to make it smaller, ideally removing the least useful phrases first. We propose a phrase table pruning method that formulates the task as a submodular function maximization problem, and solves it by using a greedy heuristic algorithm. The proposed method can scale with input size and long phrases, and experiments show that it achieves higher BLEU scores than state-of-the-art pruning methods.

There are different definitions of what a troll is. Certainly, a troll can be somebody who teases people to make them angry, or somebody who offends people, or somebody who wants to dominate any single discussion, or somebody who tries to manipulate people’s opinion (sometimes for money), etc. The last definition is the one that dominates the public discourse in Bulgaria and Eastern Europe, and this is our focus in this paper. In our work, we examine two types of opinion manipulation trolls: paid trolls that have been revealed from leaked “reputation management contracts” and “mentioned trolls” that have been called such by several different people. We show that these definitions are sensible: we build two classifiers that can distinguish a post by such a paid troll from one by a non-troll with 81-82% accuracy; the same classifier achieves 81-82% accuracy on so called mentioned troll vs. non-troll posts.

Classification of temporal textual data sequences is a common task in various domains such as social media and the Web. In this paper we propose to use Hawkes Processes for classifying sequences of temporal textual data, which exploit both temporal and textual information. Our experiments on rumour stance classification on four Twitter datasets show the importance of using the temporal information of tweets along with the textual content.

This paper studies the effect of limited precision data representation and computation on word embeddings. We present a systematic evaluation of word embeddings with limited memory and discuss method-s that directly train the limited precision representation with limited memory. Our results show that it is possible to use and train an 8 -bit fixed-point value for word embedding without loss of performance in word/phrase similarity and dependency parsing tasks.

Uncovering thematic structures of SNS and blog posts is a crucial yet challenging task, because of the severe data sparsity induced by the short length of texts and diverse use of vocabulary. This hinders effective topic inference of traditional LDA because it infers topics based on document-level co-occurrence of words. To robustly infer topics in such contexts, we propose a latent concept topic model (LCTM). Unlike LDA, LCTM reveals topics via co-occurrence of latent concepts , which we introduce as latent variables to capture conceptual similarity of words. More specifically, LCTM models each topic as a distribution over the latent concepts , where each latent concept is a localized Gaussian distribution over the word embedding space. Since the number of unique concepts in a corpus is often much smaller than the number of unique words, LCTM is less susceptible to the data sparsity. Experiments on the 20Newsgroups show the effectiveness of LCTM in dealing with short texts as well as the capability of the model in handling held-out documents with a high degree of OOV words.

We describe the Iraq Body Count Corpus (IBC-C) dataset, the first substantial armed conflict-related dataset which can be used for conflict analysis. IBC-C provides a ground-truth dataset for con-flict specific named entity recognition, slot filling, and event de-duplication. IBC-C is constructed using data collected by the Iraq Body Count project which has been recording incidents from the ongoing war in Iraq since 2003. We describe the dataset’s creation, how it can be used for the above three tasks and provide initial baseline results for the first task (named entity recognition) using Hidden Markov Models, Conditional Random Fields, and Recursive Neural Networks.

Traditional event detection methods heavily rely on manually engineered rich features. Recent deep learning approaches alleviate this problem by automatic feature engineering. But such efforts, like tradition methods, have so far only focused on single-token event mentions, whereas in practice events can also be a phrase. We instead use forward-backward recurrent neural networks (FBRNNs) to detect events that can be either words or phrases. To the best our knowledge, this is one of the first efforts to handle multi-word events and also the first attempt to use RNNs for event detection. Experimental results demonstrate that FBRNN is competitive with the state-of-the-art methods on the ACE 2005 and the Rich ERE 2015 event detection tasks.

Bilingual models that capture the semantics of sentences are typically only evaluated on cross-lingual transfer tasks such as cross-lingual document categorization or machine translation. In this work, we evaluate the quality of the monolingual representations learned with a variant of the bilingual compositional model of Hermann and Blunsom (2014), when viewing translations in a second language as a semantic annotation as the original language text. We show that compositional objectives based on phrase translation pairs outperform compositional objectives based on bilingual sentences and on monolingual paraphrases.

Neural Machine Translation (MT) has reached state-of-the-art results. However, one of the main challenges that neural MT still faces is dealing with very large vocabularies and morphologically rich languages. In this paper, we propose a neural MT system using character-based embeddings in combination with convolutional and highway layers to replace the standard lookup-based word representations. The resulting unlimited-vocabulary and affix-aware source word embeddings are tested in a state-of-the-art neural MT based on an attention-based bidirectional recurrent neural network. The proposed MT scheme provides improved results even when the source language is not morphologically rich. Improvements up to 3 BLEU points are obtained in the German-English WMT task.