Text Conversion Research Articles

Meshwork Ecology: Climate and Colonialism in Diene's Hell Freezes Over and Ruga's Public Service Announcement

ABSTRACT: Within global academia, activism, and cultural production there has been a turn towards what I conceptualize as meshwork ecology . Here, the relationality and interdependency of all earthbound things are emphasized, as is the multiplicity of human and nonhuman ontologies and epistemologies existing on the planet. This paper demonstrates how reading for meshwork ecological elements can bring to light new features of ecocentric African speculative fiction. After unpacking my understanding of meshwork ecology, I read for certain meshwork ecological elements the novella "Hell Freezes Over" by Mame Bougouma Diene and the short film Public Service Announcement by Athipatra Ruga. I show how reading for meshwork ecological elements illuminates four central aspects of these texts: 1) the interrogation of colonial ways of inhabiting the earth, 2) the centering of interconnectedness and relationality with the nonhuman, 3) thinking with multiple temporalities, and 4) exploring assemblages that involve technology, humanity, and the nonhuman. I expound on how I understand these texts in conversation with meshwork ecological scholarship as examples of planetary resonances, in this case a shared engagement with the looming reality of climate change and the pressing need to rethink how humanity relates to the nonhuman world.

Capturing multiple emotions from conversational data using fine tuned transformers

Emotion detection is one of the crucial topics of Natural Language Processing (NLP) in recent years, and now one of the biggest motivating factors in correct identification and interpretation of a wide range of emotional expressions in textual data. This research examines the use of Bidirectional Encoder Representations from Transformers (BERT), a trained transformer model for emotion detection in textual data. This analysis evaluates how good BERT is at identifying emotions such as surprise, anger, fear, happiness and sadness compared with ordinary machine learning as well as deep learning techniques. The addition of weighted emotions approach enhances the model performance and gives a deeper emotional context awareness making it more effective to deal with complicated emotional utterances. The purpose of the research is to develop a fine-tuned BERT model with a weighted emotion framework to enhance the accuracy of emotion classification in conversational text. The context is improving emotion recognition in scenarios where multiple emotions co-exist, addressing limitations in traditional models by capturing subtle and overlapping emotional expressions. In terms of training the methods multiple datasets are considered and also the research examines various models performance. The article further discusses possible application areas of BERT modified in light of NLP.

Age Group Differences in Social Media Language Use: Identity Representation via Emojis and Abbreviations

This paper aims to analyse how age groups impact the use of social media language (SL) in online textual conversations, focusing on the type, polarity, and subjectivity of SL to explore identity representation across different age groups. During data collection in 2023, there were 4.76 billion social media users worldwide; hence, understanding the use of SL is essential for analysing how individuals express their emotions and identities online. SL consists of emoticons/emojis, abbreviations, and mixed language in textual conversations. For this study, a qualitative approach was used, with 46 participants from three age groups: 25-34, 35-44, and 45-54 engaging in mock group conversations for one hour using WhatsApp. The participants were separated into six group conversations. The findings show that the older age groups (35-44 and 45-54) prefer to discuss technical or professional topics, whereas the younger age groups (25-34) are more interested in games, food, entertainment, and fun. Furthermore, for the age group 25-34 there is a high usage of emoticons/emojis, whereas the age group 45-54 uses them the least. These findings indicate that age groups have a significant influence on the use of SL and topic preferences in textual conversations, providing insights that will be used to develop a Natural Language Processing (NLP) tool for online identity classification. This study enhances human-computer interaction by investigating how different age groups use SL in digital environments.

An Emotional Dialogue System Using Conditional Generative Adversarial Networks with a Sequence-to-Sequence Transformer Encoder

Understanding the expression of emotion and generating appropriate responses are key steps toward constructing emotional, conversational agents. In this paper, we propose a framework for single-turn emotional conversation generation, and there are three main components in our model, namely, a sequence-to-sequence model with stacked encoders, a conditional variational autoencoder, and conditional generative adversarial networks. For the sequence-to-sequence model with stacked encoders, we designed a two-layer encoder by combining Transformer with gated recurrent units-based neural networks. Because of the flexibility of the sequence-to-sequence model, we adopted a conditional variational autoencoder in our framework, which uses latent variables to learn a distribution over potential responses and generates diverse responses. Furthermore, we regard a conditional variational autoencoder-based, sequence-to-sequence model as the generative model, and the training of the generative model is assisted by both a content discriminator and an emotion classifier, which assists our model in promoting content information and emotion expression. We use automated evaluation and human evaluation to evaluate our model and baselines on the NTCIR short text conversation task (STC-3) Chinese emotional conversation generation (CECG) Subtask dataset [44], and the experimental results demonstrate that our proposed framework can generate semantically reasonable and emotionally appropriate responses.

Community-Driven Research with People Who Use Drugs: A Virtual Project During Multiple Epidemics

Sociological approaches to digital and community-engaged research experienced significant innovation in recent years. This article examines developing and implementing a primarily virtual community-driven research (CDR) project with the National Survivors Union, the American national drug-users union, during the COVID-19 pandemic. Relationships between researchers and directly impacted people, such as people who use drugs, face many barriers. These issues were exacerbated during COVID-19 when in-person research decreased while drug-related harms increased. In response, this project modified the CDR model for drug-use research. The CDR model is particularly beneficial for studies with marginalized populations who may mistrust researchers. In CDR, impacted community members are fundamental project drivers. This project’s data are based on 29 months of weekly group meetings in National Survivors Union online spaces, group and individual text conversations, phone calls, and shared-document group work. The project co-developed methods for CDR with directly impacted people, including community-initiated research questions, low-threshold methods, collaborative writing strategies, coauthorship practices foregrounding directly impacted perspectives, and multiple dissemination forms. Modified CDR expands sociological methods for digital research, citizen science, and community-engaged research with vulnerable, criminalized groups. This approach may aid inclusive, innovative sociological scholarship and effective public health policy for reducing morbidity and mortality during multiple crises.

KEMAMPUAN LITERASI BAHASA INDONESIA PELAKU UMKM PRODUK LOKAL BADUI PADA TOKO ONLINE DAN E-COMMERCE TOKOPEDIA

The indigenous Badui community is known for its strict customs; among other things, prohibitions on schools and the use of technology in daily life so it is not surprising that the majority of Badui people cannot write, especially the Badui Dalam indigenous community and the older generation of the Badui Luar community. As time goes by, the Badui Luar indigenous people are starting to openup using technology in the form of cell phones and the internet. The use of cell phones and the internet among the Badui Luar indigenous community is not only for communication, but also for trading. Initially, the use of social media by the Badui indigenous community was Facebook and Instagram for trading. In the last year, some of them have started using e-commerce. The use of social media indirectly becomes a learning medium for the indigenous Badui Luar community to write because like it or not they have to communicate in writing with consumers in the form of chat texts and create product descriptions in the online shops they own. However, due to their limited ability to write and speak Indonesian well, they often experience communication barriers with consumers which ultimately impacts the success of buying and selling transactions. The objectives to be achieved in this research include: (1) knowing the ability of small business owner local Badui to write sentences, (2) knowing the ability of small business owner in understanding sentences or short conversation texts and responding to chats in writing. This research is a development of the research results of Saleh Abas (2006) which states that a person's ability to write is determined by accuracy in using languageelements, organizing discourse in the form of an essay, accuracy in using language, and choosing the words used to write. The method used in this research is descriptive qualitative. The research subjects were local Badui product who already had online shops on Tokopedia e-commerce. The sample obtained by purposive sampling consisted of 10 small business owner who already had an online shop on Tokopedia e-commerce with the status still active. Research data was obtained based on 26 texts describing local Badui products; including woven cloth, koja bags, palm sugar and tutu coffee. This data is to determine sentence structure, diction, punctuation, and creativity in creating sentences. The data analysis technique uses descriptive analysis based on product description text. The results of the research show that 6 out of 10 small business owner who are local Badui products have quite good writing skills, but are not good in terms of grammar, diction, punctuation and creativity in creating sentences. The conclusions from this research show that the Indonesian language literacy skills of small business owner local Badui product are quite good even though they have not received formal education.

A SeNA-based study of in-service teachers’ interaction patterns in an online learning community

In-service teachers, constrained by work commitments, adopt online learning methods for their graduate studies, emphasizing the significance of exploring the construction of online learning communities. Analyzing collaborative conversation texts within the online learning process of the “Learning Science” course at a university in Beijing, this study employed social epistemic network analysis (SeNA) to investigate interaction patterns of in-service teachers and propose optimization strategies. The findings of this study show that in asynchronous discussions without intervention, the overall situation of collaborative conversations among in-service teachers tends to be low-level. The social interaction network shows polycentricity, categorizing members into core and peripheral interactors based on distinct social interaction characteristics, highlighting substantial disparities in epistemic networks. Core interactors prioritize engaging with others, yet their contributions primarily echo interactions, reflecting superficial knowledge dimensions. Conversely, specific peripheral interactors prioritize independent reflection, sharing content largely derived from personal experiences or knowledge, signifying a commitment to independent and in-depth thinking within deeper knowledge dimensions. Moreover, learners’ online interactions are influenced by individual characteristics, instructional evaluation and the learning community atmosphere.

Lettrism as Implicit Philosophy in 10th-century Jewish and Muslim Creation Theories

Abstract This article presents a comparative reading of two 10th-century Islamicate lettrist works: Kitāb Khawāṣṣ al-ḥurūf by Ibn Masarra and a commentary on the Sefer Yetsirah by Saʿadia Gaon. Close analysis of both treatises emphasizes that philosophy coexists with other ways of knowing in lettrist theories. While some lettrist authors overtly employ a philosophical structure, this paper argues that others reference and re-interpret philosophical concepts (such as primordial elements and prime matter) implicitly within their work. Through reading these texts in conversation, I suggest that a framework of “implicit philosophy” creates space for lettrism within the domain of philosophy and science, while also acknowledging the intertwining aspects of mysticism, esoterism, and occultism that weave both within and beyond the realm of philosophy. Thinking in terms of implicit philosophy opens up new ways of understanding and contextualizing lettrist works, both as individual texts and as constituents of a broader Islamicate lettrist milieu.

Grammatical structures of emoji in Japanese-language text conversations

Emojis have become a ubiquitous part of everyday text communication worldwide. Cohn et al. (Cognit Res Princ Implic 4(1):1–18, 2019) studied the grammatical structure of emoji usage among English speakers and found a correlation between the sequence of emojis used and English word order, tending towards an subject–verb–object (SVO) sequence. However, it remains unclear whether emoji usage follows a universal grammar or whether it is influenced by native language grammar. Therefore, this study explored the potential influence of Japanese grammar on emoji usage by Japanese speakers. Twenty adults, all native Japanese speakers, participated in pairs. In Experiment 1, participants engaged in conversations through Google Hangouts on iPads. The experiment consisted of four conversation rounds of approximately 8 min each. The first two rounds involved one participant using only written Japanese and the other using only emojis and punctuation, with roles reversed in the second round. The third round required both participants to use only emojis and punctuation. The results indicated that participants preferred subject–object–verb (SOV) or object–verb (OV) sequences, with OV patterns being more common. This pattern reflects a distinctive attribute of Japanese grammatical structure, marked by the frequent omission of the subject. Experiment 2 substituted emojis for words, showing nouns were more commonly replaced than verbs due to the difficulty in conveying complex meanings. Reduced subject replacements again emphasised Japanese grammatical structure. In essence, emoji usage reflects native language structures, but complexities are challenging to convey, resulting in simplified sequences. This study offers insights for enhancing emoji-based communication and interface design, with implications for translation and broader communication.

Sign language video to text conversion via optimised LSTM with improved motion estimation

ABSTRACT A method of communication for the deaf and the dumb is sign language (SL). Knowing SL allows speakers and listeners to converse effectively with deaf and dumb people. Yet, while they can acquire sign language (SL) to communicate with the dumb and the deaf, untrained persons are unable to speak with them. For such individuals, an SL to text system is more beneficial in enabling them to communicate more effectively with everyday people. To interact with the deaf and dumb, SL is a tactile gesture that employs the hands and eyes. This proposes a new model for translating SL video to text conversion or generation. Frame Conversion is the initial step, where the given input video is converted into frames. Subsequently, features considered are hierarchy of skeleton and Improved Motion estimation (ME). For generating the text, this work uses an optimised LSTM model, where training of weights is done by hybrid Combined Pelican and BES algorithm (CP-BES). Moreover, the mean value for dataset ISL-CSLTR has achieved the proposed method is 27.3%, 31.6%, 28.7%, 33.4%, 29.8%, 24.5%, 36.2%, and 34.8% greater than the conventional schemes like CNN, FSU-CNN, ASL – CNN, RNN, LSTM +POA, LSTM +ARCHOA, LSTM +HGS, LSTM +PRO and LSTM +BES.

Transfer learning for emotion detection in conversational text: a hybrid deep learning approach with pre-trained embeddings

Transfer learning for emotion detection in conversational text: a hybrid deep learning approach with pre-trained embeddings

Treating young adult cannabis use disorder with text message-delivered peer network counseling

BackgroundApproximately 16.5% of U.S. young adults have a cannabis use disorder (CUD) and are at risk for negative outcomes. Treatment can reduce cannabis use, but young adults are less likely to seek help than older adults. Peer Network Counseling-txt (PNC-txt) is a brief, text-delivered, Motivational Interviewing-informed substance use intervention focusing on peer relations and activity spaces as mechanisms for behavioral change. PNC-txt has shown evidence of reducing tobacco and cannabis use with adolescents and young adults, but it has not been tested in the context of legal cannabis use. The current randomized controlled trial sought to expand the evidence regarding the context of PNC-txt effects, comparing one state in which cannabis is legal (Colorado) and one state in which it is not (Tennessee). We hypothesized that participants randomized to PNC-txt would show significant reductions in cannabis use compared to controls, with larger reductions for females and those in Colorado, and that peer relations and activity space would mediate effects. MethodsOne thousand, seventy eight 18–25 year olds (CO: 551; TN: 527) who met screening criteria for CUD and biologically-verified cannabis use were randomly assigned to PNC-txt or waitlist control condition. Every other day for 4 weeks, participants assigned to PNC-txt received pre-programmed text conversations, tailored via data from the baseline assessment. Self-report and biological indicators of cannabis use were measured at 1-, 3-, and 6-months. DiscussionData analysis is underway. Results will provide evidence regarding whether, and how, PNC-txt reduces cannabis use in young adults with CUD. Trial registrationThis trial was prospectively registered on September 28, 2020 with ClinicalTrials.gov (NCT04567394).

An action research to develop critical thinking skills in the context of citizenship education in higher education

This article aims to demonstrate how critical thinking skills can be developed as a citizenship competence in higher education. Critical thinking skills are necessary for citizens to participate in democratic processes effectively and consciously. The participants of this “action research” based study are the researcher who conducts the undergraduate course Contemporary World Issues and 30 preservice social studies teachers taking this course. Action research was conducted through distance education at a university during the COVID-19 pandemic. Data were collected through Zoom lessons, student products, semi-structured interviews, semi-structured focus group interviews, and WhatsApp text conversations. Descriptive and content analyses were conducted using MAXQDA 2020 software. According to the research findings, a curriculum based on distinguishing arguments, writing pedagogy, and online group discussions improved students’ critical thinking skills. Students used distinguishing arguments as a cognitive framework for solving social and political problems and thus developed citizenship competence. Throughout the process, feedback was found to be quite functional. The curriculum used in the action research process can be recommended for use in the development of critical thinking skills as a citizenship competency in higher education.

A Content Analysis Of Character Education In “English For Nusantara” Textbook For Seventh Grade Students

The objective of this research is to find out whether character education in the "English for Nusantara" textbook for seventh-grade students is in accordance with the current curriculum. This research analyzed character education contained in the reading and conversational texts of the textbook. The research methodology used a descriptive qualitative approach and content analysis technique. The characters found in the textbook were classified into six dimensions of the Pancasila Student Profile. The results showed that the six dimensions of the Pancasila Student Profile were found. The dimension of having faith, fear of God Almighty, having a noble character 33.2%, global diversity 12.9%, mutual cooperation 16.7%, independent 20.4%, critical reasoning 7.5%, and creative 9.3%. In the six dimensions of the Pancasila Student Profile, which have 20 elements, only 19 were found, while one element that was not found was intercultural communication and interaction. The characters in this book are mostly integrated implicitly. The primary focus dimension of the “English for the Nusantara” textbook was having faith, fear of God Almighty, and having noble character. In addition, the main focus element in this textbook was the manner towards others. Based on the classification of character education stated by Prismarani (2014), this textbook scored 95% and was classified as ‘very high’ quality.

(040) EVALUATING CHATGPT AS A SOURCE OF PATIENT EDUCATION MATERIALS ON DYSPAREUNIA

Abstract Introduction Dyspareunia, or female sexual pain, has an estimated prevalence of up to 18% worldwide. It is a complex condition with various physical and psychological etiologies, making it a challenge for patients to understand. Physicians spend significant time on patient education and providing educational materials to patients. ChatGPT is an artificial intelligence (AI) language model that generates human-like conversational text. Objective The aim of our study is to evaluate the understandability, quality, and readability of written patient educational material on topics related to dyspareunia generated by ChatGPT in English. Methods Five existing and accessible patient education handouts related to the topic of dyspareunia were identified and used as reference material. We prompted ChatGPT to create patient education leaflets on dyspareunia, female sexual dysfunction, vaginismus, and vulvodynia. The DISCERN instrument and Patient Education Materials Assessment (PEMAT) were used to assess each leaflet’s quality, understandability, and actionability. Readability of the patient education materials was assessed with the Flesch-Kincaid Grade (FKG) level. Results ChatGPT created four different patient education materials. The document created on “dyspareunia” scored 90% on understandability 75% on actionability, 42 on DISCERN, and 14.4 FKG level compared to the UpToDate (88%, 50%, 50, and 7.6) and the International Urogynecologic Association (IUGA) (100%, 0%, 80, and 13.6) on the same subject. The ChatGPT handout on “female sexual dysfunction” scored 70% on understandability, 50% on actionability, 37 on DISCERN, and 12.9 FKG Level compared to the American Urogynecologic Society (AUGS) (90%, 50%, 52, and 13) handout on the same subject. The ChatGPT handout on “vaginismus” scored 81% on understandability, 25% on actionability, 41 on DISCERN, and 11.3 FKG Level compared to the International Society for the Study of Vulvovaginal Disease (ISSVD) (90%, 0%, 44, 11.4) handout on the same subject. The ChatGPT handout on “vulvodynia” scored 81% on understandability, 25% on actionability, 36 on DISCERN, and 11.5 FKG Level compared to the ISSVD (90%, 25%, 46, 10.5) handout on the same subject. Conclusions The average PEMAT understandability scores suggest that the reference handouts (91.6%) are more understandable than the AI-generated materials (80.5%). The ChatGPT-generated handouts also fell short in quality according to the average DISCERN scores (39 vs. 54.4). The average FKG level for the AI-generated information was 12.525 compared to 11.22 for the reference patient education materials. Overall, ChatGPT was not able to create patient education handouts on topics related to dyspareunia that were as understandable or readable as the existing handouts. The ChatGPT-generated handouts did, however, score higher in actionability on average (50) than the reference handouts (25), which may give patients a false sense of trust in these AI-generated materials. Our study indicates that ChatGPT is not yet an ideal tool for the creation of patient education materials, nor is it a reliable information source for patients. Disclosure No.

National Implementation of an Artificial Intelligence-Based Virtual Dietitian for Patients With Cancer.

Nutritional status is an established driver of cancer outcomes, but there is an insufficient workforce of registered dietitians to meet patient needs for nutritional counseling. Artificial intelligence (AI) and machine learning (ML) afford the opportunity to expand access to guideline-based nutritional support. An AI-based nutrition assistant called Ina was developed on the basis of a learning data set of >100,000 expert-curated interventions, peer-reviewed literature, and clinical guidelines, and provides a conversational text message-based patient interface to guide dietary habits and answer questions. Ina was implemented nationally in partnership with 25 advocacy organizations. Data on demographics, patient-reported outcomes, and utilization were systematically collected. Between July 2019 and August 2023, 3,310 users from all 50 states registered to use Ina. Users were 73% female; median age was 57 (range, 18-91) years; most common cancer types were genitourinary (22%), breast (21%), gynecologic (19%), GI (14%), and lung (12%). Users were medically complex, with 50% reporting Stage III to IV disease, 37% with metastases, and 50% with 2+ chronic conditions. Nutritional challenges were highly prevalent: 58% had overweight/obese BMIs, 83% reported barriers to good nutrition, and 42% had food allergies/intolerances. Levels of engagement were high: 68% texted questions to Ina; 79% completed surveys; median user retention was 8.8 months; 94% were satisfied with the platform; and 98% found the guidance helpful. In an evaluation of outcomes, 84% used the advice to guide diet; 47% used recommended recipes, 82% felt the program improved quality of life (QoL), and 88% reported improved symptom management. Implementation of an evidence-based AI virtual dietitian is feasible and is reported by patients to be beneficial on diet, QoL, and symptom management. Ongoing evaluations are assessing impact on other outcomes.

Studying the Impact Of D.P.SL Model on Online Identity Management

This paper investigates online identities through social media language use, with a focus on classifying online identity within textual conversations. It sheds light on how Demographic (D) groups and Personality (P) impact the use of Social-Media Language (SL) for identity representation online. This study was conducted in 2023 when there were 4.76 billion social media users worldwide, making it essential to study how social media language is used in textual conversations to convey online identity. The study defines social media language as consisting of emoticons/emojis, abbreviations, and mixed language within textual conversations, which have become essential for expressing feelings and emotions during conversations. The D.P.SL based survey conducted for this work aimed to understand how demographic groups and personality are related to social media language. Based on the total number of social media user worldwide, 400 responses (required based on Cochran and Yamane’s formulae sample-size calculation) and the survey was distributed across various verified online survey exchange platforms. However, 406 responses were recorded with young people in age groups of 18-24 and 25-34 using social media language more as it has become a part of their social media habits. The study also found that emoticons/emojis and slang abbreviations with letter reduplication were quite common, making conversations lively and funny. Additionally, individuals whose primary language is not English use their native language but type in English for quick communication. Subsequent study is to be conducted using online mock group conversations between participating respondents to further understand correlations, causation, and concurrency on how ‘online identity’ is managed during online communications via social media language, its context of use, and polarity sentiment.

SignoSpeak: Bridging the Gap

Abstract: "SignoSpeak: Bridging the Gap" is an innovative software research aimed at transforming communication for the hearing-impaired. It converts sign language to text in real-time, breaking down communication barriers effectively. Besides text conversion, SignoSpeak interprets various gestures, enabling users to express their thoughts vividly. By translating hand and body movements into text and audio, it enriches communication and information conveyance. This device acts as a crucial link between sign language and written language, benefiting those dependent on sign language. Additionally, its ability to recognize gestures fosters inclusive cross-cultural communication. SignoSpeak signifies a significant milestone in promoting inclusivity and meaningful interaction for the hearing-impaired. Its potential lies in revolutionizing global communication norms and understanding the unique needs of individuals with hearing impairments.

Voice based E-Mail System

Abstract: The development of computer systems has made communication much simpler. Many communication features, like texting and voice calling, are available on a mobile device. We have put forth a system that will benefit people with physical disabilities. This utility allows voice to text conversion as well as text to voice conversion. With the help of this technology, we won't be able to access anything at all using keyboards—we'll only be able to utilise speech and mouse clicks. This system is also accessible to the average individual for reading purposes. It is easy to use and effective as well.

Towards bridging the semantic gap between image and text: An empirical approach

Vlogs, Recordings, news, sport coverages are huge sources of multimodal information that do not just limit to text but extend to audio, images and videos. Applications such as summary generation, image/video captioning, multimodal sentiment analysis, cross modal retrieval requires Computer Vision along with Natural Language Processing techniques to extract relevant information. Information from different modalities must be leveraged in order to extract quality content. Hence, reducing the gap between different modalities is of utmost importance. Image to text conversion is an emerging field and employs the use of encoder decoder architecture. Deep CNNs extract the feature of images and sequence to sequence models are used to generate text description. This paper is a contribution towards the growing body of research in multimodal information retrieval. In order to generate the textual description of images, we have performed 5 experiments using the benchmark Flickr8k dataset. In these experiments we have utilized different architectures - simple sequence to sequence model, attention mechanism, transformer-based architecture to name a few. The results have been evaluated using BLEAU score. Results show that the best descriptions are attained by making use of transformer architecture. We have also compared our results with the pretrained visual model vit-gpt2 that incorporates visual transformer.