Textual Cues Research Articles

Multimodal Web Page Segmentation Using Self-organized Multi-objective Clustering

Web page segmentation (WPS) aims to break a web page into different segments with coherent intra- and inter-semantics. By evidencing the morpho-dispositional semantics of a web page, WPS has traditionally been used to demarcate informative from non-informative content, but it has also evidenced its key role within the context of non-linear access to web information for visually impaired people. For that purpose, a great deal of ad hoc solutions have been proposed that rely on visual, logical, and/or text cues. However, such methodologies highly depend on manually tuned heuristics and are parameter-dependent. To overcome these drawbacks, principled frameworks have been proposed that provide the theoretical bases to achieve optimal solutions. However, existing methodologies only combine few discriminant features and do not define strategies to automatically select the optimal number of segments. In this article, we present a multi-objective clustering technique called MCS that relies on \( K \) -means, in which (1) visual, logical, and text cues are all combined in a early fusion manner and (2) an evolutionary process automatically discovers the optimal number of clusters (segments) as well as the correct positioning of seeds. As such, our proposal is parameter-free, combines many different modalities, does not depend on manually tuned heuristics, and can be run on any web page without any constraint. An exhaustive evaluation over two different tasks, where (1) the number of segments must be discovered or (2) the number of clusters is fixed with respect to the task at hand, shows that MCS drastically improves over most competitive and up-to-date algorithms for a wide variety of external and internal validation indices. In particular, results clearly evidence the impact of the visual and logical modalities towards segmentation performance.

Impacts of Cues on Learning and Attention in Immersive 360-Degree Video: An Eye-Tracking Study.

Immersive 360-degree video has become a new learning resource because of its immersive sensory experience. This study examined the effects of textual and visual cues on learning and attention in immersive 360-degree video by using eye-tracking equipment integrated in a virtual reality head-mounted display. Participants (n = 110) were randomly assigned to one of four conditions: (1) no cues, (2) textual cues in the initial field of view (FOV), (3) textual cues outside the initial FOV, and (4) textual cues outside the initial FOV + visual cues. The results showed that the cues (annotations or annotations + arrows) helped learners achieve better learning outcomes and spend more time focusing on the areas with cues. In addition, the study found a serious imbalance in the distribution of learners’ attention in each region of the video. The attention directed to textual cues in the initial FOV is much higher than the attention directed to textual cues outside the initial FOV. Adding visual cues can effectively direct attention to textual cues outside the initial FOV and alleviate the imbalance of attention distribution. Consequently, adding cues to immersive 360-degree video can be an appropriate approach to promote learning and guide attention in immersive 360-degree video learning environments. This study provided new insights into the design and development of immersive 360-degree video instructional resources.

The Impact of Product Tagging on Trust and Purchase Intention: A cross-cultural perspective in visual e-commerce

The present study aims to examine the opportunity of in-app shopping, more specifically, analyzing the influence of product tags and cultural background on consumers’ trust and purchase intention. For this purpose, a comparison was made between European and South American female consumers from two cultures: the Netherlands and Paraguay. A total of 225 subjects, 143 from the Dutch background and 82 from the Paraguayan background, participated in an experimental survey where they judged an Instagram product page (product tag: present vs. absent). The analysis of the results revealed that the Paraguayans had higher purchase intentions than the Dutch. Additionally, the presence of a text appeal in the visual product presentation had a direct positive effect on consumers’ purchase intentions and perceived trust in the shopping environment, for both the Dutch and Paraguayan cultural groups. In virtual web shops, perceiving trust is a decisive point for purchase intentions. This study contributes to the fast-growing investigations on social media effectiveness and visual marketing as an informative and persuasive tool. The findings pinpoint the synergic value of visual and textual cues of product presentation online in the atmospheric trust of in-app shopping.

Metaphors and metonymies used in memes to depict COVID-19 in Jordanian social media websites

This study provides an analysis of the monomodal and multimodal metaphors and metonymies depicting COVID-19 in a corpus of 250 memes. The theoretical framework adopted in this study included Conceptual Metaphor Theory (CMT) by Lakoff and Johnson (1980), Forceville's (2008) Mono-modal and Multimodal Metaphor Theory, and Musolff’s (2006) approach in identifying source domains scenarios. Various source domains that depict different aspects of COVID-19 were used, and some of them reflected certain aspects of the Jordanian culture. The analysis also revealed that multimodal metaphors were more frequent compared to monomodal ones. Primarily, the researchers ascribed the salient presence of multimodality to the medium-determined specificities characteristic where pictorial and textual cues are carefully selected to reinforce the message the meme intends to convey. Furthermore, conceptualizing a new pandemic required the pervasive use of multimodality. The analysis also demonstrated the crucial role of metonymy in interpreting certain conceptual metaphors.

SENTIMENT ANALYSIS FOR DEPRESSION DETECTION

The Covid-19 pandemic has dramatically changed the way we have used to live. The pandemic has been causing significant devastations in economy, and health, inter alia. Mental health, especially, has become a growing concern due to employment terminations, income loss, family stress and other uncertainties. The pandemic disproportionally affected mental health of younger population. Nowadays risk of early death is increasing due to mental illness which is mostly caused due to depression. Depression creates suicidal thoughts causing serious impairments in daily life. Sentiment analysis is a hot topic that’s been on research for decades, which intends to find the nature of text and classifies into positive, negative and neutral. In today’s digital world lot of data can be made available for sentiment analysis. Hence, our aim is to focus on creating a depression detection system from text, video & audio analysis. Sentiment Analysis and Natural Language Processing methods will be used to develop this system. The system will classify text, audio and video cues as positive or negative depending on the emotions inferred from user’s input.

The Great Refusal in Philip Roth’s American Pastoral

Ever since the publication of Philip Roth’s American Pastoral, critics have debated whether the text’s unfavorable and reductive depiction of the radicals in the long 1960s can be attributed to Roth’s own conservatism. The debate has been in part further aroused by his supposedly more sympathetic account of the fall of the protagonist “the Swede,” who embraces all things considered as standing for America itself. However, by drawing on various historical contexts and materials ―the history of the long 60s, Roth’s interviews, textual cues, and above all Herbert Marcuse’s concept of the Great Refusal―I argue that the novel is more accepting of radicalism than it allows itself and the critics allow it to be. The novel partakes in the making of the Great Refusal, that is, through which it searches for a more resilient form of radicalism. In that sense, the novel looks back at the long 60s and depicts the age with nuance, rejecting the caricature that it is crazed and enraged.

Using Visual Grouping to Improve Navigation of Skip Patterns in Mail Surveys: An Experiment

AbstractSkip patterns, in which certain questions should only be answered by subsets of respondents, are particularly challenging for mail surveys because respondents must navigate them themselves with no assistance from a computer or interviewer. Incorrect navigation of skip patterns can lead to (1) errors of omission, which is when respondents fail to answer follow-up questions they should answer; (2) errors of commission, which is when respondents answer follow-up questions they should skip; or (3) filter item nonresponse, which is when respondents fail to answer the initial filter question. Using experimental data from two general population surveys of Nebraskans, this paper examines three visual design elements—common region, indentation, and subnumbering—aimed at decreasing skip errors in mail surveys. Each design element is intended to create stronger grouping and subgrouping among items within skip patterns, thus clarifying the navigational path and decreasing skip errors. We compare error rates across treatments overall and by respondent age, education, and literacy to assess whether visual cues can provide additional aid to those who might struggle with reading text cues. We find that using separate enclosures or subnumbering of the follow-up questions significantly reduces commission error rates and using indentation significantly increases omission error rates. In addition, none of the tested visual designs were more effective for older, less educated, or lower literacy respondents than for their counterparts.

Blanche and Stanley, polar opposites. A pragmastylistic analysis of interactions from Tennessee Williams’ A Streetcar Named Desire

The paper is an attempt at demonstrating how the language used by fictional dramatic characters contributes to their characterization, that is, how the readers (audiences) perceive them based on inferences drawn from a variety of textual cues. These cues include explicit selfand other-presentation as well as implicit hints retrieved from conversation structure, aspects of turn-taking or features of the language used by the character. In this paper, Blanche DuBois and Stanley Kowalski from Tennessee Williams’ play The Streetcar Named Desire are analyzed and characterized as being polar opposites.

A framework for visual question answering with the integration of scene-text using PHOCs and fisher vectors

Text contained in an image gives useful information about that image. Consider a warning signboard with text “high voltage”; it indicates the hazard or risk involved in the image. Thus, this semantic textual information can be very useful for better understanding of images, which is not utilized by the existing visual question answering (VQA) models. However, the presence of this textual information in images can strongly guide the VQA task. This work deal with the task of visual question answering by exploiting these textual cues together with the visual content to boost the accuracy of VQA models. In the work, a novel VQA model is proposed based on the PHOC and fisher vector based representation. Based on the PHOCs of the scene-text, we have constructed a powerful descriptor by using a Fisher Vectors. Also, the proposed model uses transformer model together with dynamic pointer networks for answer decoding process. Thus, the proposed model uses a sequence of decoding steps for answer generation instead of just assuming answer prediction as a classification problem as considered by previous works. We have shown the qualitative and quantitative results on three popular datasets: VQA 2.0, TextVQA and ST-VQA. The results show the effectiveness of the proposed model over the existing models.

Examining the direct and interaction effects of picture color cues and textual cues related to color on accommodation-sharing platform rental purchase

Given the increasing popularity of peer-to-peer (P2P) accommodation-sharing platforms (i.e., Airbnb), this paper aims to explore the effects of both picture color cues and textual cues related to color (TCC) on rental decisions made on these platforms in China. Based on cue utilization theory, this paper distinguishes picture color cues signaled from photos listed by hosts and TCC signaled from online guest reviews. The paper conducts an empirical study using data crawled from Xiaozhu.com. The research model is tested with poisson regression using 2520 valid listings collected in Guangzhou and is verified with additional data collected in Shanghai. Results show that both picture color cues and TCC affect rental transactions. Properties displayed with warm-colored pictures are more likely to be rented than those with cold-colored pictures. In addition, the consistency of picture color cues and TCC is found to have a significant impact on rental decisions. Interestingly, we also found an invert-U effect of brightness on rental purchase. • Based on cue utilization theory, this paper defined picture color cues and textual cues related to color (TCC). • Results show that both picture color cues and TCC affect rental transactions. • Properties displayed with warm-colored pictures are more likely to be rented than those with cold-colored pictures. • The consistency of picture color cues and TCC is found to have a significant impact on rental decisions. • We also found an invert-U effect of brightness on rental purchase.

LITERAL BODIES (SOMATA): A TELESTICH IN OVID (METAMORPHOSES 1.406–11)

ABSTRACTThis article draws attention to the presence of a previously unnoticed transliterated telestich (SOMATA) in the transformation of stones into bodies in the episode of Deucalion and Pyrrha in Ovid's Metamorphoses (1.406–11). Detection of the Greek intext, which befits the episode's amplified bilingual atmosphere, is encouraged by a number of textual cues. The article also suggests a ludic connection to Aratus’ Phaenomena.

Distinguishing between fake news and satire with transformers

Indiscriminate elimination of harmful fake news risks destroying satirical news, which can be benign or even beneficial, because both types of news share highly similar textual cues. In this work we applied a recent development in neural network architecture, transformers, to the task of separating satirical news from fake news. Transformers have hitherto not been applied to this specific problem. Our evaluation results on a publicly available and carefully curated dataset show that the performance from a classifier framework built around a DistilBERT architecture performed better than existing machine-learning approaches. Additional improvement over baseline DistilBERT was achieved through the use of non-standard tokenization schemes as well as varying the pre-training and text pre-processing strategies. The improvement over existing approaches stands at 0.0429 (5.2%) in F1 and 0.0522 (6.4%) in accuracy. Further evaluation on two additional datasets shows our framework’s ability to generalize across datasets without diminished performance.

Beyond visual semantics: Exploring the role of scene text in image understanding

Images with visual and scene text content are ubiquitous in everyday life. However, current image interpretation systems are mostly limited to using only the visual features, neglecting to leverage the scene text content. In this paper, we propose to jointly use scene text and visual channels for robust semantic interpretation of images. We do not only extract and encode visual and scene text cues, but also model their interplay to generate a contextual joint embedding with richer semantics. The contextual embedding thus generated is applied to retrieval and classification tasks on multimedia images, with scene text content, to demonstrate its effectiveness. In the retrieval framework, we augment our learned text-visual semantic representation with scene text cues, to mitigate vocabulary misses that may have occurred during the semantic embedding. To deal with irrelevant or erroneous recognition of scene text, we also apply query-based attention to our text channel. We show how the multi-channel approach, involving visual semantics and scene text, improves upon state of the art.

Corpus Assisted Approach To Armah’s The Beautyful Ones Are Not Yet Born

Armah’s The Beautyful Ones are not yet Born is a novel known for its extensive portrayal of the ills and anomalies in the Ghanaian society right after independence. The majority of studies on the novel have overwhelmingly concluded that corruption is the preoccupation of the text. This view appears skewed in many respects. This paper argues that the corpus assisted approach can contribute methodologies to support objective investigation of the subject matters of the text. This study, adopting the corpus-assisted approach in a mix of numerical data and qualitative description of Armah’s The Beautyful Ones are not yet Born, used frequencies of the occurrence of pejorative terms in the text to determine the dominant subject matters in the novel. The approach reveals that “rot” and “decay” are the most dominant motifs used, followed by “filth”, “corruption”, and “bribery”. It suggests that clusters, i.e., recurrence of words, characters’ association with the words, and context of use serve as textual cues in thematic exploration. The approach aids in revealing that the real intent of The Beautyful Ones Are Not Yet Born is that the total breakdown of the society rests on seemingly insignificant characters. The paper has implications for methodological approaches to thematic analysis of literary texts, particularly, the novel.

Using Self-Management and Visual Cues to Improve Responses to Nonverbal Social Cues in Adults With Autism Spectrum Disorder.

Many individuals with Autism Spectrum Disorder (ASD) experience challenges with social communication, including recognizing and responding to non-verbal cues. The purpose of this study was to assess the efficacy of self-management combined with textual cues to teach adults with ASD to recognize and respond to nonverbal expressions of boredom and confusion during social conversation. A multiple baseline across participants design was used to assess the efficacy of this intervention for three participants. Results showed substantial gains across all participants in their recognition and responsiveness to the targeted nonverbal cues. Moreover, this skill maintained after the completion of intervention and generalized to novel conversation partners and settings with large effect sizes. The findings add to the literature base on interventions for adults with ASD, and further support the use of self-management and textual cues as effective intervention strategies for improving nonverbal communication.

Automatic analysis of social media images to identify disaster type and infer appropriate emergency response

Social media postings are increasingly being used in modern days disaster management. Along with the textual information, the contexts and cues inherent in the images posted on social media play an important role in identifying appropriate emergency responses to a particular disaster. In this paper, we proposed a disaster taxonomy of emergency response and used the same taxonomy with an emergency response pipeline together with deep-learning-based image classification and object identification algorithms to automate the emergency response decision-making process. We used the card sorting method to validate the completeness and correctness of the disaster taxonomy. We also used VGG-16 and You Only Look Once (YOLO) algorithms to analyze disaster-related images and identify disaster types and relevant cues (such as objects that appeared in those images). Furthermore, using decision tables and applied analytic hierarchy processes (AHP), we aligned the intermediate outputs to map a disaster-related image into the disaster taxonomy and determine an appropriate type of emergency response for a given disaster. The proposed approach has been validated using Earthquake, Hurricane, and Typhoon as use cases. The results show that 96% of images were categorized correctly on disaster taxonomy using YOLOv4. The accuracy can be further improved using an incremental training approach. Due to the use of cloud-based deep learning algorithms in image analysis, our approach can potentially be useful to real-time crisis management. The algorithms along with the proposed emergency response pipeline can be further enhanced with other spatiotemporal features extracted from multimedia information posted on social media.

Linguistic drivers of misinformation diffusion on social media during the COVID-19 pandemic

In the wake of the COVID-19 pandemic, unprecedent amounts of fake news and hoax spread on social media. In particular, conspiracy theories argued on the effect of specific new technologies like 5G and misinformation tarnished the reputation of brands like Huawei. Language plays a crucial role in understanding the motivational determinants of social media users in sharing misinformation, as people extract meaning from information based on their discursive resources and their skillset. In this paper, we analyze textual and non-textual cues from a panel of 4923 tweets containing the hashtags #5G and #Huawei during the first week of May 2020, when several countries were still adopting lockdown measures, to determine whether or not a tweet is retweeted and, if so, how much it is retweeted. Overall, through traditional logistic regression and machine learning, we found different effects of the textual and non-textual cues on the retweeting of a tweet and on its ability to accumulate retweets. In particular, the presence of misinformation plays an interesting role in spreading the tweet on the network. More importantly, the relative influence of the cues suggests that Twitter users actually read a tweet but not necessarily they understand or critically evaluate it before deciding to share it on the social media platform.

You Want This Job? Influence and Interplay of Self-Generated Text and Picture Cues in Professional Networking Service Profiles on Expertise Evaluation

ABSTRACT Job applicants’ self-presentation in their profiles on professional networking services (PNS), such as LinkedIn, may be crucial for the evaluation of their expertise for a job. Importantly, on PNS profiles, self-generated picture-based and text-based cues are shown together, and both may influence expertise evaluations. In three experiments, we systematically analyzed impression formation in the context of PNS profiles by investigating the influence of self-generated textual and pictorial cues on job applicants’ expertise evaluations. Student participants were presented with different PNS profiles that were systematically varied by the level of expertise (high vs. medium vs. low) conveyed through text (i.e., a description of experience and education) and pictures (i.e., a photo of the profile owner vs. a placeholder). Our results revealed a textual primacy in expertise evaluations of PNS profiles, as text expertise was crucial, independent of picture expertise. However, picture expertise was particularly important in cases of high text expertise. Placeholders always resulted in more negative judgments than high-expertise pictures and sometimes even had the same effect as low-expertise pictures. Finally, we discuss implications for theory building on impression formation and practical consequences for self-presentation in PNS.

Interpretable Multi-Head Self-Attention Architecture for Sarcasm Detection in Social Media.

With the online presence of more than half the world population, social media plays a very important role in the lives of individuals as well as businesses alike. Social media enables businesses to advertise their products, build brand value, and reach out to their customers. To leverage these social media platforms, it is important for businesses to process customer feedback in the form of posts and tweets. Sentiment analysis is the process of identifying the emotion, either positive, negative or neutral, associated with these social media texts. The presence of sarcasm in texts is the main hindrance in the performance of sentiment analysis. Sarcasm is a linguistic expression often used to communicate the opposite of what is said, usually something that is very unpleasant, with an intention to insult or ridicule. Inherent ambiguity in sarcastic expressions make sarcasm detection very difficult. In this work, we focus on detecting sarcasm in textual conversations from various social networking platforms and online media. To this end, we develop an interpretable deep learning model using multi-head self-attention and gated recurrent units. The multi-head self-attention module aids in identifying crucial sarcastic cue-words from the input, and the recurrent units learn long-range dependencies between these cue-words to better classify the input text. We show the effectiveness of our approach by achieving state-of-the-art results on multiple datasets from social networking platforms and online media. Models trained using our proposed approach are easily interpretable and enable identifying sarcastic cues in the input text which contribute to the final classification score. We visualize the learned attention weights on a few sample input texts to showcase the effectiveness and interpretability of our model.

How Regional Trust Cues Could Drive Decentralisation in the Energy Sector—An Exploratory Approach

The Green Information Systems research stream was initiated by leading information systems researchers to address climate change through information and communications technology. This paper responds to a call for practical research into the design of information systems that support consumers in their decision making in favour of sustainable products. We apply an exploratory approach to improve understanding of regional trust cues in the energy sector and how these could drive the sector’s decentralisation. We explore the still emerging phenomenon of regional text and imagery on digital user interfaces via a multi-method process including quantitative and qualitative content analysis. Our findings suggest that regional energy providers systematically employ regional textual and pictorial trust cues on their websites. We further lay the ground for future experimental work on this matter by defining terms and concepts and systematically capturing design elements. We outline practical implications for designing user interfaces in the energy sector and discuss how this could drive the sector’s platformisation and sustainabilisation. In addition, we discuss implications for consumers who could become the target of regional washing attempts, in other words providers applying regional cues to create a regional company image in the absence of actual regionality.