Sentence Polarity Research Articles

SENTIMENT ANALYSIS OF CODE-MIXED LANGUAGES

In the multilingual context of today, code-mixed language is frequently utilized. It happens when a sentence contains both foreign language vocabulary and grammar. Finding the sentence's polarity value is the aim of sentiment analysis of code-mixed language. It is mostly concerned with sentiment analysis of tweets that contain extra Hindi and English words and symbols. The collection is composed of 20,000 tweets, which produces word-level representations of the tweets for use as input in several models, including CNN, LSTM, and Bi-LSTM. When compared to other models, the Bi-LSTM model performs better. The precision of CNN, LSTM and Bi-LSTM is 65.00%, 58.59% and 54.24% respectively.

Looks at what isn't there: eye movements on a blank screen when processing negation in a first and a second language.

Is negation more difficult to process than affirmation? If it is, does processing negation in a second language (L2) compound the difficulty compared to the first language (L1)? This article addresses the issues of difficulties in processing different types of negation in the L1 and L2 by looking at the differences in the ways in which comprehenders anticipate upcoming visual information during sentence processing. Using a blank-screen paradigm, we recorded eye fixations of Croatian native speakers and Croatian learners of English while they were anticipating matching or mismatching pictures to sentences with various types of negation in L1 (Croatian) and L2 (English). Using a between-group design, we manipulated sentence polarity (affirmative vs. negative), negation type (sentential vs. negative quantifier) within both L1 Croatian and L2 English so that we could observe potential anticipation effects varying as a function of the two predictors. In line with previous studies, affirmation in the L1 was easier to process than negation, and participants were able to anticipate sentence-picture matches in both the L1 and the L2 group. In contrast with our prediction, anticipatory looks did not significantly vary across negation types in Croatian based on the number of structural cues. In L2 English, learners exhibited prediction ability across negation types. These findings go against the view that comprehension in L2 comes with a reduced ability to generate expectations, and they highlight the robustness of mental simulations in both L1 and L2 negation processing.

An adversarial training method for text classification

Text classification is an emerging topic in the field of text data mining, but the current methods of deducing sentence polarity have two major shortcomings: on the one hand, there is currently a lack of a large and well-curated corpus; on the other hand, current solutions based on deep learning are particularly vulnerable to attacks from adversarial samples. To overcome the limitations above, we propose an adversarial training method HNN-GRAT (Hierarchical Neural Network and Gradient Reversal) for text classification. Firstly, A Robustly Optimized BERT Pretraining Approach (RoBERTa) pretraining model is used to extract text features and feature gradient information; secondly, the original gradient information is passed through the gradient reversal layer designed to obtain the inverted gradient information; finally, the original gradient information and the inverted gradient information are fused to obtain the new gradient of the model. HNN-GRAT method are tested on three real datasets and five attack methods, compared with RoBERTa pretraining model, HNN-GRAT improves the robustness accuracy and reduces the probability of the model being attacked. In addition, using six text defense methods, HNN-GRAT achieves the best Boa and Succ (Such as, DeepWordBug attack, for AGNEWS, IMDB, SST-2 datasets, with an improvement of Boa up to 41.50%, 67.50%, 28.15% and Succ drop to 55.90%, 27.45%, 69.89%, respectively).

Activation levels of plausible alternatives in conversational negation.

Negation is often used to contradict or correct (e.g. There is no dog here.). While rejecting some state of affairs that is presumed to hold for the recipient (e.g. There is a dog here.), the speaker might implicitly suggest a set of plausible alternatives (e.g. There is a wolf instead.). Prior work indicates that alternatives are highly relevant to the comprehension of sentences involving focus: in priming studies, listeners infer plausible alternatives to focused items even when they are not contextually available. So far it is unclear whether negation similarly activates an automatic search for plausible alternatives. The current study was designed to investigate this question, by looking at the activation levels of nouns after negative and affirmative sentences. In a series of priming experiments, subjects were presented with negative and affirmative sentences (e.g. There is an/no apple.), followed by a lexical decision task with targets including plausible alternatives (e.g. pear), as well as semantically related but implausible alternatives (e.g. seed). An interaction of Sentence Polarity and Prime-Target Relation was expected, with negation facilitating responses to plausible alternatives. Results of the first experiment were numerically in line with the hypothesis but the interaction just missed significance level. A post hoc analysis revealed the expected significant interaction. Possible roles of sentential context and goodness of alternatives are discussed. A further experiment confirms that the goodness of alternatives is in fact critical in modulating the effect.

Effect of transcranialdirect current stimulation on the right brain temporal area on processing approach and avoidance attitudes with negation

Language describes approach/avoidance intentionality by means of attitudinal verbs (e.g., accept vs. reject). The right superior temporal sulcus (rSTS) has been shown to be recruited in processing action goals and approach intentionality in social contexts. In this study, we examine whether transcranial direct current stimulation (tDCS) of this area improves the processing of attitudinal verbs (either of approach or avoidance) in the context of affirmative and negative sentences [e.g., Julio (did not)/included meat on the grocery list]. After being subjected to tDCS, 46 participants were given sentences for passive reading. Sentences were displayed in segments with a fixed time of exposition, and a verb, either the one mentioned in the sentence or an alternative one was displayed 1,500 ms after the sentence (e.g., included vs. excluded, in the example). Participants were told to read them and then press the space bar to continue the experiment. Results showed shorter latencies for approach verbs that were either mentioned in approach sentences or the alternatives in avoidance sentences, both in affirmative and negative versions under anodal conditions compared to sham conditions. Thus, the anodal stimulation of rSTS affected the accessibility of approach verbs that were not modulated either by being mentioned or by sentence polarity. In addition, mentioned verbs had shorter reading times than the alternative ones in negative sentences in the anodal vs. sham condition. This suggests that stimulation caused an effect of negation in the activation of the mentioned verb. Implications are discussed in the context of the role of the rSTS in processing attitudinal verbs and negation to understand better approach and avoidance mediated by language in the framework of the two-step model of negation processing.

A hybrid lexicon-based and neural approach for explainable polarity detection

In this work, we propose BERT-WMAL, a hybrid model that brings together information coming from data through the recent transformer deep learning model and those obtained from a polarized lexicon. The result is a model for sentence polarity that manages to have performances comparable with those at the state-of-the-art, but with the advantage of being able to provide the end-user with an explanation regarding the most important terms involved with the provided prediction. The model has been evaluated on three polarity detection Italian dataset, i.e., SENTIPOLC, AGRITREND and ABSITA. While the first contains 7,410 tweets released for training and 2,000 for testing, the second and the third respectively include 1,000 tweets without splitting , and 2,365 reviews for training, 1,171 for testing. The use of lexicon-based information proves to be effective in terms of the F1 measure since it shows an improvement of F1 score on all the observed dataset: from 0.664 to 0.669 (i.e, 0.772%) on AGRITREND, from 0.728 to 0.734 (i.e., 0.854%) on SENTIPOLC and from 0.904 to 0.921 (i.e, 1.873%) on ABSITA. The usefulness of this model not only depends on its effectiveness in terms of the F1 measure, but also on its ability to generate predictions that are more explainable and especially convincing for the end-users. We evaluated this aspect through a user study involving four native Italian speakers, each evaluating 64 sentences with associated explanations. The results demonstrate the validity of this approach based on a combination of weights of attention extracted from the deep learning model and the linguistic knowledge stored in the WMAL lexicon. These considerations allow us to regard the approach provided in this paper as a promising starting point for further works in this research area.

Following negative search instructions: the role of visual context

AbstractOne typical use of negation is to refer to exceptions. In a visual display showing several similar items (majority) and one exception, referring to the exception by negating the majority should therefore be pragmatically felicitous. We investigated whether comprehenders are sensitive to these pragmatic aspects when processing negative sentences and having to identify the according items in the visual display. In Experiment 1, participants read affirmative and negative sentences referring to either the exception or the majority object in strongly biased displays. Additionally, unbiased displays were implemented, showing equal numbers of objects of each type. Identification times of the correct referent were shorter with the biased display independent of sentence polarity. Also, picking the exceptional item in the biased display was faster than picking a majority item, independent of sentence polarity. Thus, participants did not specifically profit from pragmatically felicitous conditions when processing negation. Critically, in the biased displays, the exceptional object was highly salient, which might have initially drawn the participant’s attention to this object, resulting in a general speed-up. Therefore, in Experiment 2, we used a biased display with reduced saliency of the exceptional object. Again negation did not result in a specific speed-up due to pragmatically correct negation use. Thus, negation does not seem to facilitate the identification of an exceptional object.

Bangla Text Sentiment Analysis Using Supervised Machine Learning with Extended Lexicon Dictionary

With the proliferation of the Internet’s social digital content, sentiment analysis (SA) has gained a wide research interest in natural language processing (NLP). A few significant research has been done in Bangla language domain because of having intricate grammatical structure on text. This paper focuses on SA in the context of Bangla language. Firstly, a specific domain-based categorical weighted lexicon data dictionary (LDD) is developed for analyzing sentiments in Bangla. This LDD is developed by applying the concepts of normalization, tokenization, and stemming to two Bangla datasets available in GitHub repository. Secondly, a novel rule—based algorithm termed as Bangla Text Sentiment Score (BTSC) is developed for detecting sentence polarity. This algorithm considers parts of speech tagger words and special characters to generate a score of a word and thus that of a sentence and a blog. The BTSC algorithm along with the LDD is applied to extract sentiments by generating scores of the two Bangla datasets. Thirdly, two feature matrices are developed by applying term frequency-inverse document frequency (tf-idf) to the two datasets, and by using the corresponding BTSC scores. Next, supervised machine learning classifiers are applied to the feature matrices. Results show that for the case of BiGram feature, support vector machine (SVM) achieves the best classification accuracy of 82.21% indicating the effectiveness of BTSC algorithm in Bangla SA.

Sentence polarity detection using stepwise greedy correlation based feature selection and random forests: An fMRI study

Cognitive state analysis or reading the brain was always an exciting field for researchers. Analysis of the human brain while a person is engaged in doing a particular task is an essential topic in the recent development of neuro-imaging studies. The introduction of new non-invasive methods like PET (Positron Emission Tomography) and fMRI (Functional Magnetic Resonance Imaging) is used to analyze the brain. At the same time, subjects are involved in doing diverse activities. This study aims to investigate the processing of affirmative and negative sentences in the brain. Using a greedy stepwise correlation-based feature selection technique and random forest classification approach, our model can classify the cognitive state in sentence polarity detection task with, on average, 95.41% accuracy. We have also analyzed the category-specific selected feature voxel set in determining the sentence polarity in the brain. Our result shows that CALC, RDLPFC, and LDLPFC are positively contributing areas feature selection. In contrast, RPPREC, RSGA, RFEF add very little to polarity check.

Classifying Brain State in Sentence Polarity Exposure: An ANN Model for fMRI Data

Classifying Brain State in Sentence Polarity Exposure: An ANN Model for fMRI Data

On the use of uno in Colombian Spanish: the role of transitivity

Abstract Our goal is to explore the intersection of two bodies of literature, namely, the one on impersonal constructions with an emphasis on uno ‘one’, and the one on the effect of transitivity and the focus of attention on the distribution of overt vs. null pronouns, where it has been shown that overt pronominal subjects are disfavored in transitive contexts as opposed to intransitive contexts. Through a variationist analysis of the expression of uno in Barranquilla, Colombia, in the PRESSEA-BARRANQUILLA corpus, we extend this line of inquiry to this impersonal pronoun and study in detail for the first time the effect of the various components of transitivity on the distribution of overt pronouns. Specifically, various transitivity parameters put forward by Hopper and Thompson are shown to correctly predict the distribution of uno, namely, number of participants and kinesis whereas sentence polarity, aspect and individuation of the object yield mixed results meriting future research.

Enhancement of SentiWordNet using contextual valence shifters

Sentence structure has a considerable impact on the sentiment polarity of a sentence. In the presence of contextual valence shifters like conjunctions, conditionals and intensifiers some parts of the sentence are more relevant to determine the sentence polarity. In this work we have used valence shifters in sentences to enhance the sentiment lexicon SentiWordNet in a given document set. They have also been used to improve the sentiment analysis at document level. In the near past, micro blogging services like Twitter have become an important data source for sentiment analysis. Tweets, being restricted to 140 characters have slangs, are grammatically incorrect, have spelling mistakes and have informal expressions. The method is aimed at noisy and unstructured data like tweets on which computationally intensive tools like dependency parsers are not very successful. Our proposed system works better on both noisy (Stanford and airlines datasets of Twitter) and structured (movie review) datasets.

Sentential negation of abstract and concrete conceptual categories: a brain decoding multivariate pattern analysis study.

We rarely use abstract and concrete concepts in isolation but rather embedded within a linguistic context. To examine the modulatory impact of the linguistic context on conceptual processing, we isolated the case of sentential negation polarity, in which an interaction occurs between the syntactic operator not and conceptual information in the negation's scope. Previous studies suggested that sentential negation of concrete action-related concepts modulates activation in the fronto-parieto-temporal action representation network. In this functional magnetic resonance imaging study, we examined the influence of negation on a wider spectrum of meanings, by factorially manipulating sentence polarity (affirmative, negative) and fine-grained abstract (mental state, emotion, mathematics) and concrete (related to mouth, hand, leg actions) conceptual categories. We adopted a multivariate pattern analysis approach, and tested the accuracy of a machine learning classifier in discriminating brain activation patterns associated to the factorial manipulation. Searchlight analysis was used to localize the discriminating patterns. Overall, the neural processing of affirmative and negative sentences with either an abstract or concrete content could be accurately predicted by means of multivariate classification. We suggest that sentential negation polarity modulates brain activation in distributed representational semantic networks, through the functional mediation of syntactic and cognitive control systems.This article is part of the theme issue 'Varieties of abstract concepts: development, use and representation in the brain'.

Sentiment classification with word localization based on weakly supervised learning with a convolutional neural network

In order to maximize the applicability of sentiment analysis results, it is necessary to not only classify the overall sentiment (positive/negative) of a given document but also to identify the main words that contribute to the classification. However, most datasets for sentiment analysis only have the sentiment label for each document or sentence. In other words, there is a lack of information about which words play an important role in sentiment classification. In this paper, we propose a method for identifying key words discriminating positive and negative sentences by using a weakly supervised learning method based on a convolutional neural network (CNN). In our model, each word is represented as a continuous-valued vector and each sentence is represented as a matrix whose rows correspond to the word vector used in the sentence. Then, the CNN model is trained using these sentence matrices as inputs and the sentiment labels as the output. Once the CNN model is trained, we implement the word attention mechanism that identifies high-contributing words to classification results with a class activation map, using the weights from the fully connected layer at the end of the learned CNN model. To verify the proposed methodology, we evaluated the classification accuracy and the rate of polarity words among high scoring words using two movie review datasets. Experimental results show that the proposed model can not only correctly classify the sentence polarity but also successfully identify the corresponding words with high polarity scores.

Sentential negation modulates inhibition in a stop-signal task. Evidence from behavioral and ERP data

Embodiment theories claim that language meaning involves sensory–motor simulation processes in the brain. A challenge for these theories, however, is to explain how abstract words, such as negations, are processed. In this article, we test the hypothesis that understanding sentential negation (e.g., You will not cut the bread) reuses the neural circuitry of response inhibition. Participants read manual action sentences with either affirmative or negative polarity, embedded in a Stop–Signal paradigm, while their EEG was recorded. The results showed that the inhibition-related N1 and P3 components were enhanced by successful inhibition. Most important, the early N1 amplitude was also modulated by sentence polarity, producing the largest values for successful inhibitions in the context of negative sentences, whereas no polarity effect was found for failing inhibition or go trials. The estimated neural sources for N1 effects revealed activations in the right inferior frontal gyrus, a typical inhibition-related area. Also, the estimated stop-signal reaction time was larger in trials with negative sentences. These results provide strong evidence that action-related negative sentences consume neural resources of response inhibition, resulting in less efficient processing in the Stop–Signal task.

A deep learning-based sports player evaluation model based on game statistics and news articles

Player evaluation is a key component of the question-answering (QA) system in sports. Since existing player evaluation methods heavily rely on game statistics, they cannot capture the qualitative impact of each player during a game, which can be exploited using news articles after the game. In this paper, we propose a deep learning-based player evaluation model by combining both quantitative game statistics and the qualitative analyses provided by news articles. Players are classified as positive or negative based on their performance during certain periods, and news articles in the same period are annotated using the player's class. Then, the relationship between news articles and the annotated polarity is investigated by a deep neural network, which can deal with the high dimensionality of the text data. Since there is no explicit polarity label for news articles, we use the change in game statistics in target periods to annotate related sentences. The proposed system is applied to a Korean professional baseball league (KBO) and it is shown to be capable of understanding the sentence polarity of news articles on player performances.

Intonation and Evidentiality in Spanish Polar Interrogatives.

Three different nuclear pitch accents can be found in Castilian Spanish polar interrogatives. In addition to the 'canonical' low-rise pattern, there are two marked interrogative contours featuring high-rise and rise-fall pitch accents. The aim of this paper is to explain how each contour contributes to the interpretation of the utterance in which they occur. I argue that this contribution is to be sought at the semantic, not at the pragmatic-illocutionary, attitudinal-level. My proposal is that the low-rise contour is the expression of unspecified sentence polarity (corresponding to the interrogative operator), whereas the two marked contours add indications about the information source-that is, they encode evidential distinctions. The high-rise pattern indicates that the Self is the source of the information; the rise-fall tone indicates that Other is the source. The whole range of pragmatic interpretations that have been described in the literature can be easily accommodated into the present proposal as inferential developments of the encoded meaning together with contextual information. This view has implications for a theory of interrogatives, for the phonology of intonation and for the articulation of the semantics/pragmatics interface.

A Consensus Approach to the Sentiment Analysis Problem Driven by Support-Based IOWA Majority

In group decision making, there are many situations where the opinion of the majority of participants is critical. The scenarios could be multiple, like a number of doctors finding commonality on the diagnose of an illness or parliament members looking for consensus on an specific law being passed. In this article, we present a method that utilizes induced ordered weighted averaging (IOWA) operators to aggregate a majority opinion from a number of sentiment analysis (SA) classification systems, where the latter occupy the role usually taken by human decision-makers as typically seen in group decision situations. In this case, the numerical outputs of different SA classification methods are used as input to a specific IOWA operator that is semantically close to the fuzzy linguistic quantifier ‘most of’. The object of the aggregation will be the intensity of the previously determined sentence polarity in such a way that the results represent what the majority think. During the experimental phase, the use of the IOWA operator coupled with the linguistic quantifier ‘most’ () proved to yield superior results compared to those achieved when utilizing other techniques commonly applied when some sort of averaging is needed, such as arithmetic mean or median techniques.

Sentiment Analysis of Review Datasets Using Naïve Bayes‘ and K-NN Classifier

The advent of Web 2.0 has led to an increase in the amount of sentimental content available in the Web. Such content is often found in social media web sites in the form of movie or product reviews, user comments, testimonials, messages in discussion forums etc. Timely discovery of the sentimental or opinionated web content has a number of advantages, the most important of all being monetization. Understanding of the sentiments of human masses towards different entities and products enables better services for contextual advertisements, recommendation systems and analysis of market trends. The focus of our project is sentiment focussed web crawling framework to facilitate the quick discovery of sentimental contents of movie reviews and hotel reviews and analysis of the same. We use statistical methods to capture elements of subjective style and the sentence polarity. The paper elaborately discusses two supervised machine learning algorithms: K-Nearest Neighbour(K-NN) and Naive Bayes and compares their overall accuracy, precisions as well as recall values. It was seen that in case of movie reviews Naive Bayes gave far better results than K-NN but for hotel reviews these algorithms gave lesser, almost same accuracies.

Implicit agreeing/disagreeing intention while reading self-relevant sentences: A human fMRI study

The true intentions of humans are sometimes difficult to ascertain exclusively from explicit expressions, such as speech, gestures, or facial expressions. In this experiment, functional magnetic resonance imaging (fMRI) was used to investigate implicit intentions that were generated while a subject was reading self-relevant sentences. Short sentences, which were presented visually, consisted of self-relevant statements and a substantive verb, which indicated sentence polarity as either affirmative or negative. Each sentence was divided into the contents and the sentence ending, and the subjects were asked to respond with either agreement or disagreement after the complete sentence was presented. The overall group analysis suggested that the intention of the sentence response was found even before the reading of the complete sentences. Increased neural activation was found in the left medial prefrontal cortex (MPFC) during feelings of agreement compared to feelings of disagreement during self-relevant decision-making. In addition, according to the sentence ending, the decision of a response activated the frontopolar cortex (FPC) in the switching condition. These findings indicated that the implicit intentions of responses to the given statements were internally generated before an explicit response occurred, and, hence, intentions can be used to predict a subject’s future answer.