Lexicalized Well-Founded Grammar (LWFG) is a recently developed syntactic-semantic grammar formalism for deep language understanding, which balances expressiveness with provable learnability results. The learnability result for LWFGs assumes that the semantic composition constraints are learnable. In this paper, we show what are the properties and principles the semantic representation and grammar formalism require, in order to be able to learn these constraints from examples, and give a learning algorithm. We also introduce a LWFG parser as a deductive system, used as an inference engine during LWFG induction. An example for learning a grammar for noun compounds is given.

A usual strategy to select the final answer in factoid Question-Answering (QA) relies on redundancy. A score is given to each candidate answer as a function of its frequency of occurrence, and the final answer is selected from the set of candidates sorted in decreasing order of score. For that purpose, systems often try to group together semantically equivalent answers. However, they hold several other semantic relations, such as inclusion, which are not considered, and candidates are mostly seen independently, as competitors. Our hypothesis is that not just equivalence, but other relations between candidate answers have impact on the performance of a redundancy-based QA system. In this paper, we describe experimental studies to back up this hypothesis. Our findings show that, with relatively simple techniques to recognize relations, systems' accuracy can be improved for answers of categories Number, Date and Entity.

Consistent word selection in machine translation is currently realized by resolving word sense ambiguity through the context of a single sentence or neighboring sentences. However, consistent word selection over the whole article has yet to be achieved. Consistency over the whole article is extremely important when applying machine translation to collectively developed documents like Wikipedia. In this paper, we propose to consider constraints between words in the whole article based on their semantic relatedness and contextual distance. The proposed method is successfully implemented in both statistical and rule-based translators. We evaluate those systems by translating 100 articles in the English Wikipedia into Japanese. The results show that the ratio of appropriate word selection for common nouns increased to around 75% with our method, while it was around 55% without our method.

We argue for an alternative paradigm in evaluating machine translation quality that is strongly empirical but more accurately reflects the utility of translations, by returning to a representational foundation based on AI oriented lexical semantics, rather than the superficial flat n-gram and string representations recently dominating the field. Driven by such metrics as BLEU and WER, current SMT frequently produces unusable translations where the semantic event structure is mistranslated: who did what to whom, when, where, why, and how? We argue that it is time for a new generation of more “intelligent” automatic and semi-automatic metrics, based clearly on getting the structure right at the lexical semantics level. We show empirically that it is possible to use simple PropBank style semantic frame representations to surpass all currently widespread metrics' correlation to human adequacy judgments, including even HTER. We also show that replacing human annotators with automatic semantic role labeling still yields much of the advantage of the approach. We combine the best of both worlds: from an SMT perspective, we provide superior yet low-cost quantitative objective functions for translation quality; and yet from an AI perspective, we regain the representational transparency and clear reflection of semantic utility of structural frame-based knowledge representations.

As tweets has become a comprehensive repository of fresh information, Semantic Role Labeling (SRL) for tweets has aroused great research interests because of its center role in a wide range of tweet related studies such as fine-grained information extraction, sentiment analysis and summarization. However, the fact that a tweet is often too short and informal to provide sufficient information poses a main challenge. To tackle this challenge, we propose a new method to collectively label similar tweets. The underlying idea is to exploit similar tweets to make up for the lack of information in a tweet. Specifically, similar tweets are first grouped together by clustering. Then for each cluster a two-stage labeling is conducted: One labeler conducts SRL to get statistical information, such as the predicate/argument/role triples that occur frequently, from its highly confidently labeled results; then in the second stage, another labeler performs SRL with such statistical information to refine the results. Experimental results on a human annotated dataset show that our approach remarkably improves SRL by 3.1% F1.

Trained on the imbalanced labeled data, most classification Various semi-supervised learning methods have algorithms tend to predict test samples as the majority class been proposed recently to solve the longstanding and may ignore the minority class. Although many methods, shortage problem of manually labeled data in sentiment such as re-sampling [Chawla et al., 2002], one-class classification classification. However, most existing studies [Juszczak and Duin, 2003], and cost-sensitive assume the balance between negative and positive learning [Zhou and Liu, 2006], have been proposed to solve samples in both the labeled and unlabeled data, this issue, it is still unclear as to which method is more which may not be true in reality. In this paper, we suitable to handle the imbalanced problem in sentiment investigate a more common case of semi-supervised classification and whether the method is extendable to learning for imbalanced sentiment classification.

We call this task the rating-inference task; Traditional sentiment analysis mainly considers It determines an author's polarity evaluation within a multipoint binary classifications of reviews, but in many scale (e.g. one to five "stars"). We explore solutions for real-world sentiment classification problems, nonbinary this task in the context of product or service reviews, which review ratings are more useful. This is especially are one of the most important opinion resources and widely true when consumers wish to compare two used by costumers and companies. We observe that in many products, both of which are not negative. Previous real-world scenarios, it is important to provide numerical ratings work has addressed this problem by extracting rather than binary decisions, especially when a customer various features from the review text for learning a compares several candidate products, all of them are positive predictor. Since the same word may have different in a binary classification, to make a purchase decision, since sentiment effects when used by different reviewers customers not only need to know whether a product is good or on different products, we argue that it is necessary not, but also how good the product is. A recent study pointed to model such reviewer and product dependent effects out that many consumers are willing to pay at least 20% percent in order to predict review ratings more accurately.

Event anaphora resolution plays a critical role in discourse analysis. This paper proposes a tree kernel-based framework for event pronoun resolution. In particular, a new tree expansion scheme is introduced to automatically determine a proper parse tree structure for event pronoun resolution by considering various kinds of competitive information related with the anaphor and the antecedent candidate. Evaluation on the OntoNotes English corpus shows the appropriateness of the tree kernel-based framework and the effectiveness of competitive information for event pronoun resolution.

We present unsupervised approaches to the problem of modeling dialog acts in asynchronous conversations; i.e., conversations where participants collaborate with each other at different times. In particular, we investigate a graph-theoretic deterministic framework and two probabilistic conversation models (i.e., HMM and HMM+Mix) for modeling dialog acts in emails and forums. We train and test our conversation models on (a) temporal order and (b) graph-structural order of the datasets. Empirical evaluation suggests (i) the graph-theoretic framework that relies on lexical and structural similarity metrics is not the right model for this task, (ii) conversation models perform better on the graph-structural order than the temporal order of the datasets and (iii) HMM+Mix is a better conversation model than the simple HMM model.

In order to deal with ambiguity in natural language, it is common to organise words, according to their senses, in synsets, which are groups of synonymous words that can be seen as concepts. The manual creation of a broad-coverage synset base is a time-consuming task, so we take advantage of dictionary definitions for extracting synonymy pairs and clustering for identifying synsets. Since word senses are not discrete, we create fuzzy synsets, where each word has a membership degree. We report on the results of the creation of a fuzzy synset base for Portuguese, from three electronic dictionaries. The resulting resource is larger than existing hancrafted Portuguese thesauri.