Although many studies on sentiment analysis have been carried out for widely spoken languages, this topic is still immature for Turkish. Most of the works in this language focus on supervised models, which necessitate comprehensive annotated corpora. There are a few unsupervised methods, and they utilize sentiment lexicons either built by translating from English lexicons or created based on corpora. This results in improper word polarities as the language and domain characteristics are ignored. In this paper, we develop unsupervised (domain-independent) and semi-supervised (domain-specific) methods for Turkish, which are based on a set of antonym word pairs as seeds. We make a comprehensive analysis of supervised methods under several feature weighting schemes. We then form ensemble of supervised classifiers and also combine the unsupervised and supervised methods. Since Turkish is an agglutinative language, we perform morphological analysis and use different word forms. The methods developed were tested on two datasets having different styles in Turkish and also on datasets in English to show the portability of the approaches across languages. We observed that the combination of the unsupervised and supervised approaches outperforms the other methods, and we obtained a significant improvement over the state-of-the-art results for both Turkish and English.

This work introduces robust multi-dialectal part of speech tagging trained on an annotated data set of Arabic tweets in four major dialect groups: Egyptian, Levantine, Gulf, and Maghrebi. We implement two different sequence tagging approaches. The first uses conditional random fields (CRFs), while the second combines word- and character-based representations in a deep neural network with stacked layers of convolutional and recurrent networks with a CRF output layer. We successfully exploit a variety of features that help generalize our models, such as Brown clusters and stem templates. Also, we develop robust joint models that tag multi-dialectal tweets and outperform uni-dialectal taggers. We achieve a combined accuracy of 92.4% across all dialects, with per dialect results ranging between 90.2% and 95.4%. We obtained the results using a train/dev/test split of 70/10/20 for a data set of 350 tweets per dialect.

Subwords have become very popular, but the BERTa and ERNIEbtokenizers often produce surprising results. Byte pair encoding (BPE) trains a dictionary with a simple information theoretic criterion that sidesteps the need for special treatment of unknown words. BPE is more about training (populating a dictionary of word pieces) than inference (parsing an unknown word into word pieces). The parse at inference time can be ambiguous. Which parse should we use? For example, “electroneutral” can be parsed as electron-eu-tral or electro-neutral, and “bidirectional” can be parsed as bid-ire-ction-al and bi-directional. BERT and ERNIE tend to favor the parse with more word pieces. We propose minimizing the number of word pieces. To justify our proposal, a number of criteria will be considered: sound, meaning, etc. The prefix, bi-, has the desired vowel (unlike bid) and the desired meaning (bi is Latin for two, unlike bid, which is Germanic for offer).

The recent breakthroughs in deep neural architectures across multiple machine learning fields have led to the widespread use of deep neural models. These learners are often applied as black-box models that ignore or insufficiently utilize a wealth of preexisting semantic information. In this study, we focus on the text classification task, investigating methods for augmenting the input to deep neural networks (DNNs) with semantic information. We extract semantics for the words in the preprocessed text from the WordNet semantic graph, in the form of weighted concept terms that form a semantic frequency vector. Concepts are selected via a variety of semantic disambiguation techniques, including a basic, a part-of-speech-based, and a semantic embedding projection method. Additionally, we consider a weight propagation mechanism that exploits semantic relationships in the concept graph and conveys a spreading activation component. We enrich word2vec embeddings with the resulting semantic vector through concatenation or replacement and apply the semantically augmented word embeddings on the classification task via a DNN. Experimental results over established datasets demonstrate that our approach of semantic augmentation in the input space boosts classification performance significantly, with concatenation offering the best performance. We also note additional interesting findings produced by our approach regarding the behavior of term frequency - inverse document frequency normalization on semantic vectors, along with the radical dimensionality reduction potential with negligible performance loss.

Translations are generally assumed to share universal features that distinguish them from texts that are originally written in the same language. Thus, we can argue that these translations constitute their own variety of a language, often called translationese. However, translations are also influenced by their source languages and thus show different characteristics depending on the source language. Consequently, we argue that these variants constitute different “dialects” of translations into the same target language. Studies using machine learning techniques on Indo-European languages have investigated the universal characteristics of translationese and how translations from various source languages differ. However, for typologically very different languages such as Chinese, there are only few corpus studies that tap into the intricate relation between translations and the originals, as well as into the relations among translations themselves. In this contribution, we investigate the following questions: (1) What are the characteristics of Chinese translationese, both in general and with respect to different source languages? (2) Can we find differences not only at the lexical but also on the syntactic level? and (3) Based on the characteristics found in the previous questions, which of the proposed laws and universals can we corroborate based on our evidence from Chinese? We use machine learning to operationalize determining the importance of different characteristics and comparing their importance for our Chinese dataset with characteristics previously reported in studies on English. In addition, our methodology allows us to add syntactic features, which have rarely been used to study translations into Chinese. Our results show that Chinese translations as a whole can be reliably distinguished from non-translations, even based on only five features. More interestingly, typological traces from the source languages can often be found in their translations, therefore creating what we call dialects of translationese. For instance, translations from two Altaic languages exhibit more noun repetition and less frequent use of pronouns. Additionally, some characteristics that are not discriminative for English work well for Chinese, possibly because the distance between Chinese and the source languages is greater than that in English studies.

Arabic sentiment analysis models have recently employed compositional paragraph or sentence embedding features to represent the informal Arabic dialectal content. These embeddings are mostly composed via ordered, syntax-aware composition functions and learned within deep neural network architectures. With the differences in the syntactic structure and words’ order among the Arabic dialects, a sentiment analysis system developed for one dialect might not be efficient for the others. Here we present syntax-ignorant, sentiment-specific n-gram embeddings for sentiment analysis of several Arabic dialects. The novelty of the proposed model is illustrated through its features and architecture. In the proposed model, the sentiment is expressed by embeddings, composed via the unordered additive composition function and learned within a shallow neural architecture. To evaluate the generated embeddings, they were compared with the state-of-the art word/paragraph embeddings. This involved investigating their efficiency, as expressive sentiment features, based on the visualisation maps constructed for our n-gram embeddings and word2vec/doc2vec. In addition, using several Eastern/Western Arabic datasets of single-dialect and multi-dialectal contents, the ability of our embeddings to recognise the sentiment was investigated against word/paragraph embeddings-based models. This comparison was performed within both shallow and deep neural network architectures and with two unordered composition functions employed. The results revealed that the introduced syntax-ignorant embeddings could represent single and combinations of different dialects efficiently, as our shallow sentiment analysis model, trained with the proposed n-gram embeddings, could outperform the word2vec/doc2vec models and rival deep neural architectures consuming, remarkably, less training time.

The rise of social media empowers people to interact and communicate with anyone anywhere in the world. The possibility of being anonymous avoids censorship and enables freedom of expression. Nevertheless, this anonymity might lead to cybersecurity issues, such as opinion spam, sexual harassment, incitement to hatred or even terrorism propaganda. In such cases, there is a need to know more about the anonymous users and this could be useful in several domains beyond security and forensics such as marketing, for example. In this paper, we focus on a fine-grained analysis of language varieties while considering also the authors’ demographics. We present a Low-Dimensionality Statistical Embedding method to represent text documents. We compared the performance of this method with the best performing teams in the Author Profiling task at PAN 2017. We obtained an average accuracy of 92.08% versus 91.84% for the best performing team at PAN 2017. We also analyse the relationship of the language variety identification with the authors’ gender. Furthermore, we applied our proposed method to a more fine-grained annotated corpus of Arabic varieties covering 22 Arab countries and obtained an overall accuracy of 88.89%. We have also investigated the effect of the authors’ age and gender on the identification of the different Arabic varieties, as well as the effect of the corpus size on the performance of our method.

This paper proposes a robust text classification and correspondence analysis approach to identification of similar languages. In particular, we propose to use the readily available information of clauses and word length distribution to model similar languages. The modeling and classification are based on the hypothesis that languages are self-adaptive complex systems and hence can be classified by dynamic features describing the system, especially in terms of distributional relations of constituents of a system. For similar languages whose grammatical differences are often subtle, classification based on dynamic system features should be more effective. To test this hypothesis, we considered both regional and genre varieties of Mandarin Chinese for classification. The data are extracted from two comparable balanced corpora to minimize possible confounding factors. The two corpora are the Sinica Corpus from Taiwan and the Lancaster Corpus of Mandarin Chinese from Mainland China, and the two genres are reportage and review. Our text classification and correspondence analysis results show that the linguistically felicitous two-level constituency model combining power functions between word and clauses effectively classifies the two varieties of Chinese for both genres. In addition, we found that genres do have compounding effect on classification of regional varieties. In particular, reportage in two varieties is more likely to be classified than review, corroborating the complex system view of language variations. That is, language variations and changes typically do not take place evenly across the board for the complete language system. This further enhances our hypothesis that dynamic complex system features, such as the power functions captured by the Menzerath–Altmann law, provide effective models in classifications of similar languages.

Technical writing in professional environments, such as user manual authoring, requires the use of uniform language. Nonuniform language refers to sentences in a technical document that are intended to have the same meaning within a similar context, but use different words or writing style. Addressing this nonuniformity problem requires the performance of two tasks. The first task, which we named nonuniform language detection (NLD), is detecting such sentences. We propose an NLD method that utilizes different similarity algorithms at lexical, syntactic, semantic and pragmatic levels. Different features are extracted and integrated by applying a machine learning classification method. The second task, which we named nonuniform language correction (NLC), is deciding which sentence among the detected ones is more appropriate for that context. To address this problem, we propose an NLC method that combines contraction removal, near-synonym choice, and text readability comparison. We tested our methods using smartphone user manuals. We finally compared our methods against state-of-the-art methods in paraphrase detection (for NLD) and against expert annotators (for both NLD and NLC). The experiments demonstrate that the proposed methods achieve performance that matches expert annotators.

Anomaly detection can be seen as an unsupervised learning task in which a predictive model created on historical data is used to detect outlying instances in new data. This work addresses possibly promising but relatively uncommon application of anomaly detection to text data. Two English-language and one Polish-language Internet discussion forums devoted to psychoactive substances received from home-grown plants, such as hashish or marijuana, serve as text sources that are both realistic and possibly interesting on their own, due to potential associations with drug-related crime. The utility of two different vector text representations is examined: the simple bag of words representation and a more refined Global Vectors (GloVe) representation, which is an example of the increasingly popular word embedding approach. They are both combined with two unsupervised anomaly detection methods, based on one-class support vector machines (SVM) and based on dissimilarity to k-medoids clusters. The GloVe representation is found definitely more useful for anomaly detection, permitting better detection quality and ameliorating the curse of dimensionality issues with text clustering. The cluster dissimilarity approach combined with this representation outperforms one-class SVM with respect to detection quality and appears a more promising approach to anomaly detection in text data.

The ACL-2019 Business meeting ended with a discussion of reviewing. Conferences are experiencing a success catastrophe. They are becoming bigger and bigger, which is not only a sign of success but also a challenge (for reviewing and more). Various proposals for reducing submissions were discussed at the Business meeting. IMHO, the problem is not so much too many submissions, but rather, random reviewing. We cannot afford to do reviewing as badly as we do (because that leads to even more submissions). Negative feedback loops are effective. The reviewing process will improve over time if reviewers teach authors how to write better submissions, and authors teach reviewers how to write more constructive reviews. If you have received a not-ok (unhelpful/offensive) review, please help program committees improve by sharing your not-ok reviews on social media.

Efficiently exploiting all sources of information such as labeled instances, classes’ representation, and relations of them has a high impact on the performance of Multi-Label Text Classification (MLTC) systems. Most of the current approaches use labeled documents as the primary source of information for MLTC. We investigate the effectiveness of different sources of information— such as the labeled training data, textual labels of classes, and taxonomy relations of classes— for MLTC. More specifically, first, for each document–class pair, different features are extracted using different sources of information. The features reflect the similarity of classes and documents. Then, MLTC is considered to be a ranking problem, and a learning to rank (LTR) approach is used for ranking classes regarding documents and selecting labels of documents. An important characteristic of many MLTC instances is that documents can belong to multiple classes and there are implicit relations between classes. We apply score propagation on top of LTR to incorporate co-occurrence patterns of classes in labeled documents. Our main findings are the following. First, using an LTR approach integrating all features, we observe significantly better performance than previous systems for MLTC. Specifically, we show that simple classification approaches fail when there is a high number of classes. Second, the analysis of feature weights reveals the relative importance of various sources of evidence, also giving insight into the underlying classification problem. Interestingly, the results indicate that the titles of documents are more informative than all other sources of information. Third, a lean-and-mean system using only four features is able to perform at 96% of the large LTR model that we propose in this paper. Fourth, using the co-occurrence information of classes helps in classifying documents more accurately. Our results show that the co-occurrence information is more helpful when the underlying classifier has a poor performance.

We present the novel task of understanding multi-sentence entity-seeking questions (MSEQs), that is, the questions that may be expressed in multiple sentences, and that expect one or more entities as an answer. We formulate the problem of understanding MSEQs as a semantic labeling task over an open representation that makes minimal assumptions about schema or ontology-specific semantic vocabulary. At the core of our model, we use a BiLSTM (bidirectional LSTM) conditional random field (CRF), and to overcome the challenges of operating with low training data, we supplement it by using BERT embeddings, hand-designed features, as well as hard and soft constraints spanning multiple sentences. We find that this results in a 12–15 points gain over a vanilla BiLSTM CRF. We demonstrate the strengths of our work using the novel task of answering real-world entity-seeking questions from the tourism domain. The use of our labels helps answer 36% more questions with 35% more (relative) accuracy as compared to baselines. We also demonstrate how our framework can rapidly enable the parsing of MSEQs in an entirely new domain with small amounts of training data and little change in the semantic representation.

In this paper, we present two approaches and the implemented system for bilingual terminology extraction that rely on an aligned bilingual domain corpus, a terminology extractor for a target language, and a tool for chunk alignment. The two approaches differ in the way terminology for the source language is obtained: the first relies on an existing domain terminology lexicon, while the second one uses a term extraction tool. For both approaches, four experiments were performed with two parameters being varied. In the experiments presented in this paper, the source language was English, and the target language Serbian, and a selected domain was Library and Information Science, for which an aligned corpus exists, as well as a bilingual terminological dictionary. For term extraction, we used the FlexiTerm tool for the source language and a shallow parser for the target language, while for word alignment we used GIZA++. The evaluation results show that for the first approach the F1 score varies from 29.43% to 51.15%, while for the second it varies from 61.03% to 71.03%. On the basis of the evaluation results, we developed a binary classifier that decides whether a candidate pair, composed of aligned source and target terms, is valid. We trained and evaluated different classifiers on a list of manually labeled candidate pairs obtained after the implementation of our extraction system. The best results in a fivefold cross-validation setting were achieved with the Radial Basis Function Support Vector Machine classifier, giving a F1 score of 82.09% and accuracy of 78.49%.

In this article, we investigate using deep neural networks with different word representation techniques for named entity recognition (NER) on Turkish noisy text. We argue that valuable latent features for NER can, in fact, be learned without using any hand-crafted features and/or domain-specific resources such as gazetteers and lexicons. In this regard, we utilize character-level, character n-gram-level, morpheme-level, and orthographic character-level word representations. Since noisy data with NER annotation are scarce for Turkish, we introduce a transfer learning model in order to learn infrequent entity types as an extension to the Bi-LSTM-CRF architecture by incorporating an additional conditional random field (CRF) layer that is trained on a larger (but formal) text and a noisy text simultaneously. This allows us to learn from both formal and informal/noisy text, thus improving the performance of our model further for rarely seen entity types. We experimented on Turkish as a morphologically rich language and English as a relatively morphologically poor language. We obtained an entity-level F1 score of 67.39% on Turkish noisy data and 45.30% on English noisy data, which outperforms the current state-of-art models on noisy text. The English scores are lower compared to Turkish scores because of the intense sparsity in the data introduced by the user writing styles. The results prove that using subword information significantly contributes to learning latent features for morphologically rich languages.