Computer Science

Sentence-level relation extraction (RE) aims at identifying the relationship between two entities in a sentence. Many efforts have been devoted to this problem, while the best performing methods are still far from perfect. In this paper, we revisit two problems that affect the performance of existing RE models, namely entity representation and noisy or ill-defined labels. Our improved baseline model, incorporated with entity representations with typed markers, achieves an F1 of 74.6% on TACRED, significantly outperforms previous SOTA methods. Furthermore, the presented new baseline achieves an F1 of 91.1% on the refined Re-TACRED dataset, demonstrating that the pre-trained language models achieve unexpectedly high performance on this task. We release our code to the community for future research.

As major progress is made in open-ended text generation, measuring how close machine-generated text is to human language remains a critical open problem. We propose Mauve, a comparison measure for open-ended text generation, which directly compares a generation model's distribution to that of human-written text. Mauve measures the mean area under a divergence curve for the two distributions, exploring the trade-off between two types of errors: those arising from parts of the human distribution that the model distribution approximates well, and those it does not. Mauve extends a family of information divergence metrics, introducing a tractable approximation based on computing the KL divergence in a quantized embedding space. This yields an efficient implementation that scales up to modern text generation models. Through an extensive empirical study on three open-ended generation tasks, we find that Mauve identifies known properties of generated text, scales naturally with model size, and correlates with human judgments, with fewer restrictions than existing distributional evaluation metrics.

Text-to-SQL is a crucial task toward developing methods for understanding natural language by computers. Recent neural approaches deliver excellent performance; however, models that are difficult to interpret inhibit future developments. Hence, this study aims to provide a better approach toward the interpretation of neural models. We hypothesize that the internal behavior of models at hand becomes much easier to analyze if we identify the detailed performance of schema linking simultaneously as the additional information of the text-to-SQL performance. We provide the ground-truth annotation of schema linking information onto the Spider dataset. We demonstrate the usefulness of the annotated data and how to analyze the current state-of-the-art neural models.

Entity Linking is one of the essential tasks of information extraction and natural language understanding. Entity linking mainly consists of two tasks: recognition and disambiguation of named entities. Most studies address these two tasks separately or focus only on one of them. Moreover, most of the state-of-the -art entity linking algorithms are either supervised, which have poor performance in the absence of annotated corpora or language-dependent, which are not appropriate for multi-lingual applications. In this paper, we introduce an Unsupervised Language-Independent Entity Disambiguation (ULIED), which utilizes a novel approach to disambiguate and link named entities. Evaluation of ULIED on different English entity linking datasets as well as the only available Persian dataset illustrates that ULIED in most of the cases outperforms the state-of-the-art unsupervised multi-lingual approaches.

Natural Language Processing (NLP) is today a very active field of research and innovation. Many applications need however big sets of data for supervised learning, suitably labelled for the training purpose. This includes applications for the Arabic language and its national dialects. However, such open access labeled data sets in Arabic and its dialects are lacking in the Data Science ecosystem and this lack can be a burden to innovation and research in this field. In this work, we present an open data set of social data content in several Arabic dialects. This data was collected from the Twitter social network and consists on +50K twits in five (5) national dialects. Furthermore, this data was labeled for several applications, namely dialect detection, topic detection and sentiment analysis. We publish this data as an open access data to encourage innovation and encourage other works in the field of NLP for Arabic dialects and social media. A selection of models were built using this data set and are presented in this paper along with their performances.

While Curriculum Learning (CL) has recently gained traction in Natural language Processing Tasks, it is still not adequately analyzed. Previous works only show their effectiveness but fail short to explain and interpret the internal workings fully. In this paper, we analyze curriculum learning in sentiment analysis along multiple axes. Some of these axes have been proposed by earlier works that need more in-depth study. Such analysis requires understanding where curriculum learning works and where it does not. Our axes of analysis include Task difficulty on CL, comparing CL pacing techniques, and qualitative analysis by visualizing the movement of attention scores in the model as curriculum phases progress. We find that curriculum learning works best for difficult tasks and may even lead to a decrement in performance for tasks with higher performance without curriculum learning. We see that One-Pass curriculum strategies suffer from catastrophic forgetting and attention movement visualization within curriculum pacing. This shows that curriculum learning breaks down the challenging main task into easier sub-tasks solved sequentially.

Good communication is indubitably the foundation of effective teamwork. Over time teams develop their own communication styles and often exhibit entrainment, a conversational phenomena in which humans synchronize their linguistic choices. This paper examines the problem of predicting team performance from embeddings learned from multiparty dialogues such that teams with similar conflict scores lie close to one another in vector space. Embeddings were extracted from three types of features: 1) dialogue acts 2) sentiment polarity 3) syntactic entrainment. Although all of these features can be used to effectively predict team performance, their utility varies by the teamwork phase. We separate the dialogues of players playing a cooperative game into stages: 1) early (knowledge building) 2) middle (problem-solving) and 3) late (culmination). Unlike syntactic entrainment, both dialogue act and sentiment embeddings are effective for classifying team performance, even during the initial phase. This finding has potential ramifications for the development of conversational agents that facilitate teaming.

In zero-shot cross-lingual transfer, a supervised NLP task trained on a corpus in one language is directly applicable to another language without any additional training. A source of cross-lingual transfer can be as straightforward as lexical overlap between languages (e.g., use of the same scripts, shared subwords) that naturally forces text embeddings to occupy a similar representation space. Recently introduced cross-lingual language model (XLM) pretraining brings out neural parameter sharing in Transformer-style networks as the most important factor for the transfer. In this paper, we aim to validate the hypothetically strong cross-lingual transfer properties induced by XLM pretraining. Particularly, we take XLM-RoBERTa (XLMR) in our experiments that extend semantic textual similarity (STS), SQuAD and KorQuAD for machine reading comprehension, sentiment analysis, and alignment of sentence embeddings under various cross-lingual settings. Our results indicate that the presence of cross-lingual transfer is most pronounced in STS, sentiment analysis the next, and MRC the last. That is, the complexity of a downstream task softens the degree of crosslingual transfer. All of our results are empirically observed and measured, and we make our code and data publicly available.

Search is one of the most common platforms used to seek information. However, users mostly get overloaded with results whenever they use such a platform to resolve their queries. Nowadays, direct answers to queries are being provided as a part of the search experience. The question-answer (QA) retrieval process plays a significant role in enriching the search experience. Most off-the-shelf Semantic Textual Similarity models work fine for well-formed search queries, but their performances degrade when applied to a domain-specific setting having incomplete or grammatically ill-formed search queries in prevalence. In this paper, we discuss a framework for calculating similarities between a given input query and a set of predefined questions to retrieve the question which matches to it the most. We have used it for the financial domain, but the framework is generalized for any domain-specific search engine and can be used in other domains as well. We use Siamese network [6] over Long Short-Term Memory (LSTM) [3] models to train a classifier which generates unnormalized and normalized similarity scores for a given pair of questions. Moreover, for each of these question pairs, we calculate three other similarity scores: cosine similarity between their average word2vec embeddings [15], cosine similarity between their sentence embeddings [7] generated using RoBERTa [17] and their customized fuzzy-match score. Finally, we develop a metaclassifier using Support Vector Machines [19] for combining these five scores to detect if a given pair of questions is similar. We benchmark our model's performance against existing State Of The Art (SOTA) models on Quora Question Pairs (QQP) dataset as well as a dataset specific to the financial domain.

Aspect-based Sentiment analysis (ABSA) accomplishes a fine-grained analysis that defines the aspects of a given document or sentence and the sentiments conveyed regarding each aspect. This level of analysis is the most detailed version that is capable of exploring the nuanced viewpoints of the reviews. Most of the research available in ABSA focuses on English language with very few work available on Arabic. Most previous work in Arabic has been based on regular methods of machine learning that mainly depends on a group of rare resources and tools for analyzing and processing Arabic content such as lexicons, but the lack of those resources presents another challenge. To overcome these obstacles, Deep Learning (DL)-based methods are proposed using two models based on Gated Recurrent Units (GRU) neural networks for ABSA. The first one is a DL model that takes advantage of the representations on both words and characters via the combination of bidirectional GRU, Convolutional neural network (CNN), and Conditional Random Field (CRF) which makes up (BGRU-CNN-CRF) model to extract the main opinionated aspects (OTE). The second is an interactive attention network based on bidirectional GRU (IAN-BGRU) to identify sentiment polarity toward extracted aspects. We evaluated our models using the benchmarked Arabic hotel reviews dataset. The results indicate that the proposed methods are better than baseline research on both tasks having 38.5% enhancement in F1-score for opinion target extraction (T2) and 7.5% in accuracy for aspect-based sentiment polarity classification (T3). Obtaining F1 score of 69.44% for T2, and accuracy of 83.98% for T3.