A deep network model for paraphrase detection in short text messages

This paper is concerned with paraphrase detection, i.e., identifying sentences that are semantically identical. The ability to detect similar sentences written in natural language is crucial for several applications, such as text mining, text summarization, plagiarism detection, authorship authentication and question answering. Recognizing this importance, we study in particular how to address the challenges with detecting paraphrases in user generated short texts, such as Twitter, which often contain language irregularity and noise, and do not necessarily contain as much semantic information as longer clean texts. We propose a novel deep neural network-based approach that relies on coarse-grained sentence modelling using a convolutional neural network (CNN) and a recurrent neural network (RNN) model, combined with a specific fine-grained word-level similarity matching model. More specifically, we develop a new architecture, called DeepParaphrase, which enables to create an informative semantic representation of each sentence by (1) using CNN to extract the local region information in form of important n-grams from the sentence, and (2) applying RNN to capture the long-term dependency information. In addition, we perform a comparative study on state-of-the-art approaches within paraphrase detection. An important insight from this study is that existing paraphrase approaches perform well when applied on clean texts, but they do not necessarily deliver good performance against noisy texts, and vice versa. In contrast, our evaluation has shown that the proposed DeepParaphrase-based approach achieves good results in both types of texts, thus making it more robust and generic than the existing approaches.     

What the fake? Probing misinformation detection standing on the shoulder of novelty and emotion

One of the most time-critical challenges for the Natural Language Processing (NLP) community is to combat the spread of fake news and misinformation. Existing approaches for misinformation detection use neural network models, statistical methods, linguistic traits, fact-checking strategies, etc. However, the menace of fake news seems to grow more vigorous with the advent of humongous and unusually creative language models. Relevant literature reveals that one major characteristic of the virality of fake news is the presence of an element of surprise in the story, which attracts immediate attention and invokes strong emotional stimulus in the reader. In this work, we leverage this idea and propose textual novelty detection and emotion prediction as the two tasks relating to automatic misinformation detection. We re-purpose textual entailment for novelty detection and use the models trained on large-scale datasets of entailment and emotion to classify fake information. Our results correlate with the idea as we achieve state-of-the-art (SOTA) performance (7.92%, 1.54%, 17.31% and 8.13% improvement in terms of accuracy) on four large-scale misinformation datasets. We hope that our current probe will motivate the community to explore further research on misinformation detection along this line. The source code is available at the GitHub.²     

Quality flaw prediction in Spanish Wikipedia: A case of study with verifiability flaws

In this work, we present the first quality flaw prediction study for articles containing the two most frequent verifiability flaws in Spanish Wikipedia: articles which do not cite any references or sources at all (denominated Unreferenced) and articles that need additional citations for verification (so-called Refimprove). Based on the underlying characteristics of each flaw, different state-of-the-art approaches were evaluated. For articles not citing any references, a well-established rule-based approach was evaluated and interesting findings show that some of them suffer from Refimprove flaw instead. Likewise, for articles that need additional citations for verification, the well-known PU learning and one-class classification approaches were evaluated. Besides, new methods were compared and a new feature was also proposed to model this latter flaw. The results showed that new methods such as under-bagged decision trees with sum or majority voting rules, biased-SVM, and centroid-based balanced SVM, perform best in comparison with the ones previously published.     

Semi-supervised ranking aggregation

Ranking aggregation is a task of combining multiple ranking lists given by several experts or simple rankers to get a hopefully better ranking. It is applicable in several fields such as meta search and collaborative filtering. Most of the existing work is under an unsupervised framework. In these methods, the performances are usually limited especially in unreliable case since labeled information is not involved in. In this paper, we propose a semi-supervised ranking aggregation method, in which preference constraints of several item pairs are given. In our method, the aggregation function is learned based on the ordering agreement of different rankers. The ranking scores assigned by this ranking function on the labeled data should be consistent with the given pairwise order constraints while the ranking scores on the unlabeled data obey the intrinsic manifold structure of the rank items. The experimental results on toy data and the OHSUMED data are presented to illustrate the validity of our method.     

Rank-based self-training for graph convolutional networks

Graph Convolutional Networks (GCNs) have been established as a fundamental approach for representation learning on graphs, based on convolution operations on non-Euclidean domain, defined by graph-structured data. GCNs and variants have achieved state-of-the-art results on classification tasks, especially in semi-supervised learning scenarios. A central challenge in semi-supervised classification consists in how to exploit the maximum of useful information encoded in the unlabeled data. In this paper, we address this issue through a novel self-training approach for improving the accuracy of GCNs on semi-supervised classification tasks. A margin score is used through a rank-based model to identify the most confident sample predictions. Such predictions are exploited as an expanded labeled set in a second-stage training step. Our model is suitable for different GCN models. Moreover, we also propose a rank aggregation of labeled sets obtained by different GCN models. The experimental evaluation considers four GCN variations and traditional benchmarks extensively used in the literature. Significant accuracy gains were achieved for all evaluated models, reaching results comparable or superior to the state-of-the-art. The best results were achieved for rank aggregation self-training on combinations of the four GCN models.     

Semi-supervised document retrieval

This paper proposes a new machine learning method for constructing ranking models in document retrieval. The method, which is referred to as SSRank, aims to use the advantages of both the traditional Information Retrieval (IR) methods and the supervised learning methods for IR proposed recently. The advantages include the use of limited amount of labeled data and rich model representation. To do so, the method adopts a semi-supervised learning framework in ranking model construction. Specifically, given a small number of labeled documents with respect to some queries, the method effectively labels the unlabeled documents for the queries. It then uses all the labeled data to train a machine learning model (in our case, Neural Network). In the data labeling, the method also makes use of a traditional IR model (in our case, BM25). A stopping criterion based on machine learning theory is given for the data labeling process. Experimental results on three benchmark datasets and one web search dataset indicate that SSRank consistently and almost always significantly outperforms the baseline methods (unsupervised and supervised learning methods), given the same amount of labeled data. This is because SSRank can effectively leverage the use of unlabeled data in learning.     

基于改进的K-means算法的异常检测

聚类算法通常用于数据的聚类。除此,它还可以用于异常数据的检测。首先介绍了基于划分的聚类算法K-means,然后给出改进算法I-K-means的算法描述,最后通过实例进行异常分析。     

Neural entity alignment with cross-modal supervision

The majority of currently available entity alignment (EA) solutions primarily rely on structural information to align entities, which is biased and disregards additional multi-source information. To compensate for inadequate structural details, this article suggests the SKEA framework, which is a simple but flexible framework for Entity Alignment with cross-modal supervision of Supporting Knowledge. We employ a relational aggregate network to specifically utilize the details about the entity and its neighbors. To overcome the limitations of relational features, two multi-modal encode modules are being used to extract visual and textural information. A new set of potential aligned entity pairs are generated by SKEA in each iteration using the knowledge of two reference modalities, which can enhance the model’s supervision. It is important to note that the supporting information used in our framework does not participate in the network’s backpropagation, which considerably improves efficiency and differs dramatically from earlier work. In comparison to existing baselines, experiments demonstrate that our proposed framework can incorporate multi-aspect information efficiently and enable supervisory signals from other modalities to transmit to entities. The maximum performance improvement of 5.24% indicates our suggested framework’s superiority, especially for sparse KGs.     

Towards effective document clustering: A constrained K-means based approach

Document clustering is an important tool for document collection organization and browsing. In real applications, some limited knowledge about cluster membership of a small number of documents is often available, such as some pairs of documents belonging to the same cluster. This kind of prior knowledge can be served as constraints for the clustering process. We integrate the constraints into the trace formulation of the sum of square Euclidean distance function of K-means. Then,the combined criterion function is transformed into trace maximization, which is further optimized by eigen-decomposition. Our experimental evaluation shows that the proposed semi-supervised clustering method can achieve better performance, compared to three existing methods.     

A Recursive tree-structured neural network with goal forgetting and information aggregation for solving math word problems

Most existing state-of-the-art neural network models for math word problems use the Goal-driven Tree-Structured decoder (GTS) to generate expression trees. However, we found that GTS does not provide good predictions for longer expressions, mainly because it does not capture the relationships among the goal vectors of each node in the expression tree and ignores the position order of the nodes before and after the operator. In this paper, we propose a novel Recursive tree-structured neural network with Goal Forgetting and information aggregation (RGFNet) to address these limits. The goal forgetting and information aggregation module is based on ordinary differential equations (ODEs) and we use it to build a sub-goal information feedback neural network (SGIFNet). Unlike GTS, which uses two-layer gated-feedforward networks to generate goal vectors, we introduce a novel sub-goal generation module. The sub-goal generation module could capture the relationship among the related nodes (e.g. parent nodes, sibling nodes) using attention mechanism. Experimental results on two large public datasets i.e. Math23K and Ape-clean show that our tree-structured model outperforms the state-of-the-art models and obtains answer accuracy over 86%. Furthermore, the performance on long-expression problems is promising.¹