Recognition of Fine-Grained Emotions from Text: An Approach Based on the Compositionality Principle

In this paper, we address the tasks of recognition and interpretation of affect communicated through text messaging in online communication environments. Specifically, we focus on Instant Messaging (IM) or blogs, where people use an informal or garbled style of writing. We introduced a novel rule-based linguistic approach for affect recognition from text. Our Affect Analysis Model (AAM) was designed to deal with not only grammatically and syntactically correct textual input, but also informal messages written in an abbreviated or expressive manner. The proposed rule-based approach processes each sentence in stages, including symbolic cue processing, detection and transformation of abbreviations, sentence parsing and word/phrase/sentence-level analyses. Our method is capable of processing sentences of different complexity, including simple, compound, complex (with complement and relative clauses) and complex–compound sentences. Affect in text is classified into nine emotion categories (or neutral). The strength of the resulting emotional state depends on vectors of emotional words, relations among them, tense of the analysed sentence and availability of first person pronouns. The evaluation of the Affect Analysis Model algorithm showed promising results regarding its capability to accurately recognize fine-grained emotions reflected in sentences from diary-like blog posts (averaged accuracy is up to 77 per cent), fairy tales (averaged accuracy is up to 70.2 per cent) and news headlines (our algorithm outperformed eight other systems on several measures).

The recognition of personal emotional state or sentiment conveyed through text is the main task we address in our research. The communication of emotions through text messaging and posts of personal blogs poses the ‘informal style of writing’ challenge for researchers expecting grammatically correct input. Our Affect Analysis Model was designed to handle the informal messages written in an abbreviated or expressive manner. While constructing our rule-based approach to affect recognition from text, we followed the compositionality principle. Our method is capable of processing sentences of different complexity, including simple, compound, complex (with complement and relative clauses), and complex-compound sentences. The evaluation of the Affect Analysis Model algorithm showed promising results regarding its capability to accurately recognize affective information in text from an existing corpus of personal blog posts.

User study on AffectIM, an avatar-based Instant Messaging system employing rule-based affect sensing from text

In this paper, we address the tasks of recognition and interpretation of affect communicated through text messaging. The evolving nature of language in online conversations is a main issue in affect sensing from this media type, since sentence parsing might fail while syntactical structure analysis. The developed Affect Analysis Model was designed to handle not only correctly written text, but also informal messages written in abbreviated or expressive manner. The proposed rule-based approach processes each sentence in sequential stages, including symbolic cue processing, detection and transformation of abbreviations, sentence parsing, and word/phrase/sentence-level analyses. In a study based on 160 sentences, the system result agrees with at least two out of three human annotators in 70% of the cases. In order to reflect the detected affective information and social behaviour, an avatar was created.KeywordsAffective sensing from textaffective user interfaceavataremotionsonline communicationlanguage parsing and understandingtext analysis

Frontiers Events is a rapidly growing calendar management system dedicated to the scheduling of academic events. This includes announcements and invitations, participant listings and search functionality, abstract handling and publication, related events and post-event exchanges. Whether an organizer or participant, make your event a Frontiers Event!

Event Abstract Back to Event Syntactic comprehension in aphasia. An evaluation test with relative clauses in Spanish María E. Sánchez1*, Analí Taboh1, Martin Fuchs2, Juan P. Barreyro3, 4 and Virginia Jaichenco1 1 University of Buenos Aires, Linguistics Institute, Argentina 2 Yale University, Department of Linguistics, United States 3 University of Buenos Aires, Faculty of Psychology, Argentina 4 CONICET, Argentina This work aimed to try out an instrument that can detect the sentences comprehension deficits by manipulating two types of structures with relative clauses (subject and object) in Spanish. We compare the performance between aphasic patients and their controls using a binary sentence–picture matching task. The type of structure was manipulated: El oso que patea al perro es azul [The bear that kicks the dog is blue] (subject relative clause) vs. El oso al que patea el perro es azul [The bear that the dog kicks is blue] (object relative clause). The test was administered to 151 native Spanish speakers, of 3 age groups and 3 different schooling levels, and a group of 5 aphasic patients. The results showed that in the control group there is a strong interaction between the type of sentence and the level of schooling, with more errors in the sentences with a relative object as the level of schooling decreases. Aphasic patients, as a group, did not differ from the lowest schooling groups in subject relative clauses, but did diverge from all groups in object relative clauses. Within the group of patients, we found that three patients (AG, RD and RR) did not differ from their control group in subject relative clauses, but in object relative clauses. Contrary, a patient (OV) differs significantly from their control group in the two structures, that is, it is worse in both object and subject relative clauses (although in the case of object clause, their performance is much worse). Finally, RC patient does not differ significantly in any of the two structures with their control group. The data allow establishing differences between patients with and without syntactic alterations. First, the data allow us to discuss how schooling affects the processing of these two structures in Spanish in subjects without language impairment. In addition, the evidence shows that agrammatic aphasic patients clearly have difficulties to assign the thematic roles to non-canonical structures, and this test is sensitive to detect them. References del Río, D., López-Higes, R., & Martín-Aragoneses, M. T. (2012). Canonical word order and interference-based integration costs during sentence comprehension: The case of Spanish subject and object relative clauses. Quarterly Journal of Experimental Psychology, 65, 2108–2128 Friedmann, N. (2008). Traceless relatives: Agrammatic comprehension of relative clauses with resumptive pronouns. Journal of Neurolinguistics, 21(2), 138-149. Garraffa, M. & Grillo, N. (2008). Canonicity effects as grammatical phenomena. Journal of Neurolinguistics, 21, 177-197. Grodzinsky, Y. (1989). Agrammatic comprehension of relative clauses. Brain and Language, 37, 480–499. Keywords: sentence comprehension, relative clauses, Evaluation, agrammatism, neurolinguistics Conference: Academy of Aphasia 55th Annual Meeting , Baltimore, United States, 5 Nov - 7 Nov, 2017. Presentation Type: poster presentation Topic: General Submission Citation: Sánchez ME, Taboh A, Fuchs M, Barreyro JP and Jaichenco V (2019). Syntactic comprehension in aphasia. An evaluation test with relative clauses in Spanish. Conference Abstract: Academy of Aphasia 55th Annual Meeting . doi: 10.3389/conf.fnhum.2017.223.00071 Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters. The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated. Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed. For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions. Received: 27 Apr 2017; Published Online: 25 Jan 2019. * Correspondence: PhD. María E Sánchez, University of Buenos Aires, Linguistics Institute, Buenos Aires, Argentina, mariaelinasanchez@yahoo.com.ar Login Required This action requires you to be registered with Frontiers and logged in. To register or login click here. Abstract Info Abstract The Authors in Frontiers María E Sánchez Analí Taboh Martin Fuchs Juan P Barreyro Virginia Jaichenco Google María E Sánchez Analí Taboh Martin Fuchs Juan P Barreyro Virginia Jaichenco Google Scholar María E Sánchez Analí Taboh Martin Fuchs Juan P Barreyro Virginia Jaichenco PubMed María E Sánchez Analí Taboh Martin Fuchs Juan P Barreyro Virginia Jaichenco Related Article in Frontiers Google Scholar PubMed Abstract Close Back to top Javascript is disabled. Please enable Javascript in your browser settings in order to see all the content on this page.

Comprehension of sentences with a center-embedded, object-gapped relative clause (ORC) is challenging for children as well as adults. Mismatching lexical and grammatical features of subject noun phrases (NPs) across the main clause and relative clause has been shown to facilitate comprehension. Adani et al. concluded that children’s comprehension improved under conditions of NP number mismatch (e.g., singular main clause subject and plural relative clause subject) as compared with NP number match (e.g., both singular subjects). However, their stimuli provided number information on verb phrases (VPs) as well as NPs creating a confound for conclusions about facilitative effects of NP number mismatch. In this study, we isolated the contribution of NP number mismatch. Notably, 32 6-year-olds with typical language participated in a center-embedded, ORC sentence comprehension task with 4 types of stimuli: (a) NP number mismatch without VP number information (NP mismatch only), (b) NP number match without VP number information (NP match only), (c) NP number mismatch with VP number mismatch (NP + VP mismatch), and (d) NP number match with VP number match (NP + VP match). Children selected one of four pictures in an array to a verbally presented relative clause sentence; 56 sentences were presented. The within-subjects comparison for NP mismatch only and NP match only was not significant. However, the within-subjects comparison for NP mismatch only and NP + VP mismatch was significant. Children were more successful in NP + VP mismatch sentence comprehension ([Formula: see text] = 0.70).

The goal of this study is to examine the processing of Chinese relative clauses (RCs) through a self-paced reading task and to determine whether the learning environment plays a role in the second-language (L2) acquisition of RCs. We investigated two types of RCs (subject vs. object RCs) along with two positions in which a RC can occur (modifying a matrix subject noun phrase [NP] vs. a matrix object NP). Eighteen native speakers of Chinese and twenty-one L2 learners at an intermediate proficiency level participated in the study. Ten learners were students learning Chinese in the US (i. e., in a foreign-language context), whereas the other eleven learners were students studying Chinese in China (i. e., in a study-abroad context). The comprehension of sentences containing a RC and reading times (RTs) on the RC and the head noun (the segment immediately following the RC) were analyzed. The results show distinct patterns for the learners and the native speakers. The accuracy data reveals that the L2 learners in China performed better than the L2 learners in the US. Additionally, the L2 learners in China exhibited a processing speed advantage to the L2 learners in the US. The RT data highlighted important asymmetries in the L2 learners in the US and the native speakers, while the results were flat for the L2 learners in China. Specifically, L2 learners in the US took longer to read object RCs than subject RCs while the opposite pattern was obtained for the L1 speakers. Moreover, matrix-object-modifying RCs revealed shorter RTs than matrix-subject-modifying RCs for L2 learners in the US, whereas the opposite pattern was found for the L1 speakers. These findings are discussed in light of the Linear Distance Theory and the Structural Distance Theory (e. g., O’Grady 1997.Syntactic development. Chicago: University of Chicago Press). Overall, these results seem to provide support to the assumption that changes in syntactic processing happen as a result of exposure to the language environment (Cuetos et al. 1996. Parsing in different languages. In Manuel Carreias, Jose E. Garcia-Albea & Nuria Sebastien-Galles (eds.),Language processing in Spanish, 145–187. Mahwah, NJ: Erlbaum; Frenck–Mestre 2002. An on-line look at sentence processing in the second language. In Roberto Heredia & Jeanette Altarriba (eds.),Bilingual sentence processing, 217–236. Amsterdam: Elsevier Science Publishers.).

The paper focuses on a novel system iFeel_IM! that integrates 3D virtual world Second Life, intelligent component for automatic emotion recognition from text messages, and innovative affective haptic interfaces providing additional nonverbal communication channels through simulation of emotional feedback and social touch (physical co-presence). The core component, Affect Analysis Model, automatically recognizes nine emotions from text. The detected emotion is stimulated by innovative haptic devices integrated into iFeel_IM!. Users can not only exchange messages but also emotionally and physically feel the presence of the communication partner (e.g., family member, friend, or beloved person).

Two experiments, one using self-paced reading and one using eye tracking, investigated the influence of noun animacy on the processing of subject relative (SR) clauses, object relative (OR) clauses, and object relative clauses with stylistic inversion (OR-SI) in French. Each sentence type was presented in two versions: either with an animate relative clause (RC) subject and an inanimate object (AS/IO), or with an inanimate RC subject and an animate object (IS/AO). There was an interaction between the RC structure and noun animacy. The advantage of SR sentences over OR and OR-SI sentences disappeared in AS/IO sentences. The interaction between animacy and structure occurred in self-paced reading times and in total fixation times on the RCs, but not in first-pass reading times. The results are consistent with a late interaction between animacy and structural processing during parsing and provide data relevant to several models of parsing.

First paragraph: MacDonald (2013) proposes that distributional properties of language and processing biases in language comprehension can to a large extent be attributed to consequences of the language production process. In essence, the account is derived from the principle of least effort that was formulated by Zipf, among others (Zipf, 1949; Levelt, 2013). However, in Zipf's view the outcome of the least effort principle was a compromise between least effort for the speaker and least effort for the listener, whereas MacDonald puts most of the burden on the production process.

The present study investigates the relationship between working memory (WM) and second language (L2) sentence processing, specifically focusing on the comprehension of English relative clauses (RCs). While previous studies have observed that object RCs typically impose greater cognitive demands than subject RCs, the role of WM in mediating this difficulty remains unclear. To investigate this relationship, we employed a self-paced reading method to assess reading times and comprehension accuracy among advanced Korean learners of English. WM was measured using both digit span and reading span tasks. Results revealed that higher WM was associated with faster reading speeds and better comprehension across RC types. Although object RCs took longer to process, there was a lack of a significant interaction between WM and RC types. Peak reading times were observed at region 4 (corresponding to the object position in subject RCs and the main verb position in object RCs). This pattern diverges from previous research findings and suggests that advanced learners engage in proactive resource allocation. These results emphasize WM’s role in L2 sentence processing and indicate processing strategies for improving learners’ management of complex syntactic structures.

The paper does concern the IM (Instant Messaging) system with a cryptographic feature designed for the portable subscriber appliances working with the Android operating system. Unlike the existing applications with a text messaging function, the proposed system uses XML (Extensible Markup Language) tool to specify the message structure and in order to ensure appropriate confidentiality of talks it does encrypt messages to be transmitted between the end user and server system. The results of a preliminary performance evaluation of encryption algorithms, used in the proposed system, will be discussed together with possible applications and further modifications of that IM system.

Emotion classification is a key task in natural language processing with lots of applications in our daily life.Traditional methods, such as Logistic Regression, struggle to capture the complex semantic and contextual dependencies in informal and diverse text, often leading to suboptimal performance. To address these limitations, we propose a novel approach that combines BERT with BiLSTM networks and utilizes data augmentation techniques. BERTs pre-trained embeddings contain rich contextual information, while the BiLSTM layer models long-range dependencies within the text. Data augmentation is applied to expand the training dataset, improving the model's generalization and robustness. Experimental results on the ACL IMDB dataset demonstrate that our approach outperforms traditional models, with the BERT + BiLSTM model achieving a testing accuracy of 88%, compared to 82% for BERT alone. This study shows that combining BERT, BiLSTM, and data augmentation significantly enhances the accuracy and robustness of emotion classification models, particularly for complex and informal text.

Informal short texts on the web are rich in emotions as they often reflect unfiltered immediate reactions to breaking news events. The emotion density, however, stands in contrast to its poverty of linguistic contexts and features for emotion classification. This paper tackles that challenge by proposing orthographic features based on orthographic code mixing and code-switching for both non-ML and ML approaches. Our results show that orthographic features routinely outperform grammatical features for emotion classification for short texts in all approaches as expected. Orthographic features were also shown to make more significant contributions, especially in terms of precision and in formal texts when state of the art deep learning algorithms are applied. This result confirms the effectiveness of the orthographic change feature to the task of emotion classification. These results are argued to be applicable to all languages because of the common code-shifting in languages with non-Latin orthographies, and the use of non-letter symbols in all languages.