Abstract | We present a graph-based semi-supervised learning for the question-answering (QA) task for ranking candidate sentences. |
Abstract | With a new representation of graph-based SSL on QA datasets using only a handful of features, and under limited amounts of labeled data, we show improvement in generalization performance over state-of-the-art QA models. |
Conclusions and Discussions | In this paper, we applied a graph-based SSL algorithm to improve the performance of QA task by exploiting unlabeled entailment relations between affirmed question and candidate sentence pairs. |
Conclusions and Discussions | We demonstrated that summarization on graph-based SSL can improve the QA task performance when more unlabeled data is used to learn the classifier model. |
Experiments | We evaluated the performance of graph-based QA system using a set of 202 questions from the TREC04 as testing dataset (Voorhees, 2003), (Prager et al., 2000). |
Graph Based Semi-Supervised Learning for Entailment Ranking | In general graph-based SSL, a function over the graph is estimated such that it satisfies two conditions: 1) close to the observed labels , and 2) be smooth on the whole graph by: |
Graph Based Semi-Supervised Learning for Entailment Ranking | Most graph-based SSLs are transductive, i.e., not easily expendable to new test points outside L U U. |
Graph Summarization | Research on graph-based SSL algorithms point out their effectiveness on real applications, e.g., (Zhu et al., 2003), (Zhou and Scholkopf, 2004), (Sindhwani et al., 2007). |
Graph Summarization | Using graph-based SSL method on the new representative dataset, X’ = X U XTe, which is comprised of summarized dataset, X = {Xifizy as labeled data points, and the testing dataset, XTe as unlabeled data points. |
Introduction | Recent research indicates that using labeled and unlabeled data in semi-supervised learning (SSL) environment, with an emphasis on graph-based methods, can improve the performance of information extraction from data for tasks such as question classification (Tri et al., 2006), web classification (Liu et al., 2006), relation extraction (Chen et al., 2006), passage-retrieval (Otterbacher et al., 2009), various natural language processing tasks such as part-of-speech tagging, and named-entity recognition (Suzuki and Isozaki, 2008), word-sense disam- |
Introduction | We construct a textual entailment (TE) module by extracting features from each paired question and answer sentence and designing a classifier with a novel yet feasible graph-based SSL method. |
Dependency Parsing | There has been much recent work on dependency parsing using graph-based , transition-based, and hybrid methods; see Nivre and McDonald (2008) for an overview. |
Dependency Parsing | Typical graph-based methods consider linear classifiers of the form |
Dependency Parsing as an ILP | Our approach will build a graph-based parser without the drawback of a restriction to local features. |
Experiments | baselines, all of them state-of-the-art parsers based on non-arc-factored models: the second order model of McDonald and Pereira (2006), the hybrid model of Nivre and McDonald (2008), which combines a (labeled) transition-based and a graph-based parser, and a refinement of the latter, due to Martins et al. |
Experiments | Comparing with the baselines, we observe that our full model outperforms that of McDonald and Pereira (2006), and is in line with the most accurate dependency parsers (Nivre and McDonald, 2008; Martins et al., 2008), obtained by combining transition-based and graph-based parsers.14 Notice that our model, compared with these hybrid parsers, has the advantage of not requiring an ensemble configuration (eliminating, for example, the need to tune two parsers). |
Experiments | 13Unlike our model, the hybrid models used here as baselines make use of the dependency labels at training time; indeed, the transition-based parser is trained to predict a labeled dependency parse tree, and the graph-based parser use these predicted labels as input features. |