Description Coding Research Articles

User Views on Online Sexual Health Symptom Checker Tool: Qualitative Research.

Delayed diagnosis and treatment of sexually transmitted infections (STIs) contributes to poorer health outcomes and onward transmission to sexual partners. Access to best-practice sexual health care may be limited by barriers such as cost, distance to care providers, sexual stigma, and trust in health care providers. Online assessments of risk offer a novel means of supporting access to evidence-based sexual health information, testing, and treatment by providing more individualized sexual health information based on user inputs. This developmental evaluation aims to find potential users' views and experiences in relation to an online assessment of risk, called iSpySTI (Melbourne Sexual Health Center), including the likely impacts of use. Individuals presenting with urogenital symptoms to a specialist sexual health clinic were given the opportunity to trial a web-based, Bayesian-powered tool that provides a list of 2 to 4 potential causes of their symptoms based on inputs of known STI risk factors and symptoms. Those who tried the tool were invited to participate in a once-off, semistructured research interview. Descriptive, action, and emotion coding informed the comparative analysis of individual cases. Findings from interviews with 14 people who had used the iSpySTI tool support the superiority of the online assessment of STI risk compared to existing sources of sexual health information (eg, internet search engines) in providing trusted and probabilistic information to users. Additionally, potential users reported benefits to their emotional well-being in the intervening period between noticing symptoms and being able to access care. Differences in current and imagined urgency of health care seeking and emotional impacts were found based on clinical diagnosis (eg, non-STI, curable and incurable but treatable STIs) and whether participants were born in Australia or elsewhere. Online assessments of risk provide users experiencing urogenital symptoms with more individualized and evidence-based health information that can improve their health care-seeking and provide reassurance in the period before they can access care.

Qudit stabilizer codes, CFTs, and topological surfaces

We study general maps from the space of rational conformal field theories (CFTs) with a fixed chiral algebra and associated Chern-Simons (CS) theories to the space of qudit stabilizer codes with a fixed generalized Pauli group. We consider certain natural constraints on such a map and show that the map can be described as a graph homomorphism from an orbifold graph, which captures the orbifold structure of CFTs, to a code graph, which captures the structure of self-dual stabilizer codes. By studying explicit examples, we show that this graph homomorphism cannot always be a graph embedding. However, we construct a physically motivated map from universal orbifold subgraphs of CFTs to operators in a generalized Pauli group. We show that this map results in a self-dual stabilizer code if and only if the surface operators in the bulk CS theories corresponding to the CFTs in question are self-dual. For CFTs admitting a stabilizer code description, we show that the full Abelianized generalized Pauli group can be obtained from twisted sectors of certain 0-form symmetries of the CFT. Finally, we connect our construction with SymTFTs, and we argue that many equivalences between codes that arise in our setup correspond to equivalence classes of bulk topological surfaces under fusion with invertible surfaces. Published by the American Physical Society 2024

Multiple description image coding using Schur transform

A novel obvious and robust image Multiple Description Coding (MDC) using Schur transform (ST)is presented in this work. To overcome the inherent drawbacks in the discrete wavelet transform (DWT)which are the hardware implementation and computational complexityIn the MDC approach, two or more correlated descriptions of the same data aregenerated, which can be independently decoded, and yield mutually refinable information. Therefore, the quality of the recovered signal is dependent only on the number of received descriptions, and not on the specific loss pattern. The new techniquewas analyzed and compared to the conventional approach in the case of four descriptions by using multiple grayscale test images having different spectral characteristics.Theexperimental results of Schur transform (ST) compared to discrete wavelet transform (DWT)show that our method achievesbetter performance in terms of delivered peaksignal to noise ratio (PSNR) and visual image quality reconstruction.The proposed MDC has a low redundancy and outperforms the conventionalscheme for image transmission over error-prone networks.Therefore, the proposed approach provides an easier practical implementation and a better redundancy performancethan the classical approach.

Service language and small talk: differentiating genuine from mundane in B2B contexts

PurposeThis study examines the role of genuine small talk in business-to-business (B2B) communication through the lens of service language. By exploring and distinguishing genuine and mundane small talk, the research elucidates the impact of sincere, empathetic interactions on value co-creation and relational dynamics within the B2B contexts.Design/methodology/approachEmploying an abductive qualitative research strategy, we conducted semi-structured interviews with 25 B2B professionals from Australian and Indonesian multinational corporations. Data analysis involved descriptive coding and thematic analysis, supplemented by network view strategies to visualize interrelations within the dataset.FindingsThe study finds that genuine small talk, characterized by sincere and contextually relevant exchanges, enhances relational enrichment, customer satisfaction and long-term loyalty in B2B interactions. Participants highlighted the importance of empathetic communication and personalized conversations in building trust and rapport. Conversely, mundane small talk, marked by superficiality, was found to erode trust and diminish the quality of client relationships. These insights underscore the critical importance of integrating genuine small talk into communication strategies to foster enduring business relationships and enhance overall service effectiveness.Originality/valueThis study contributes to the service literature by exploring the notion of genuine and mundane small talk in B2B communication, grounding them in service language perspective. It provides novel insights into the verbal and non-verbal elements of genuine small talk, offering practical implications for enhancing B2B negotiation strategies.

Possible Health Benefits and Risks of DeepFake Videos: A Qualitative Study in Nursing Students.

"DeepFakes" are synthetic performances created by AI, using neural networks to exchange faces in images and modify voices. Due to the novelty and limited literature on its risks/benefits, this paper aims to determine how young nursing students perceive DeepFake technology, its ethical implications, and its potential benefits in nursing. This qualitative study used thematic content analysis (the Braun and Clarke method) with videos recorded by 50 third-year nursing students, who answered three questions about DeepFake technology. The data were analyzed using ATLAS.ti (version 22), and the project was approved by the Ethics Committee (code UCV/2021-2022/116). Data analysis identified 21 descriptive codes, classified into four main themes: advantages, disadvantages, health applications, and ethical dilemmas. Benefits noted by students include use in diagnosis, patient accompaniment, training, and learning. Perceived risks include cyberbullying, loss of identity, and negative psychological impacts from unreal memories. Nursing students see both pros and cons in DeepFake technology and are aware of the ethical dilemmas it poses. They also identified promising healthcare applications that could enhance nurses' leadership in digital health, stressing the importance of regulation and education to fully leverage its potential.

Data-Driven Control for Nonlinear Networked Systems on Basis of Two Description Coding Scheme

Data-Driven Control for Nonlinear Networked Systems on Basis of Two Description Coding Scheme

Can GPT-3.5 generate and code discharge summaries?

The aim of this study was to investigate GPT-3.5 in generating and coding medical documents with International Classification of Diseases (ICD)-10 codes for data augmentation on low-resource labels. Employing GPT-3.5 we generated and coded 9606 discharge summaries based on lists of ICD-10 code descriptions of patients with infrequent (or generation) codes within the MIMIC-IV dataset. Combined with the baseline training set, this formed an augmented training set. Neural coding models were trained on baseline and augmented data and evaluated on an MIMIC-IV test set. We report micro- and macro-F1 scores on the full codeset, generation codes, and their families. Weak Hierarchical Confusion Matrices determined within-family and outside-of-family coding errors in the latter codesets. The coding performance of GPT-3.5 was evaluated on prompt-guided self-generated data and real MIMIC-IV data. Clinicians evaluated the clinical acceptability of the generated documents. Data augmentation results in slightly lower overall model performance but improves performance for the generation candidate codes and their families, including 1 absent from the baseline training data. Augmented models display lower out-of-family error rates. GPT-3.5 identifies ICD-10 codes by their prompted descriptions but underperforms on real data. Evaluators highlight the correctness of generated concepts while suffering in variety, supporting information, and narrative. While GPT-3.5 alone given our prompt setting is unsuitable for ICD-10 coding, it supports data augmentation for training neural models. Augmentation positively affects generation code families but mainly benefits codes with existing examples. Augmentation reduces out-of-family errors. Documents generated by GPT-3.5 state prompted concepts correctly but lack variety, and authenticity in narratives.

Sources of moral distress among obstetrician-gynecologists after Dobbs: A qualitative, multi-state study

Since the US Supreme Court's 2022 decision in Dobbs v Jackson Women's Health Organization, 18 states have enacted functional bans on abortion, yet little is known about how these laws contribute to workplace stress and its sequelae among clinicians. The purpose of this study was to characterize sources of moral distress—which occurs when a clinician knows the right course of clinical action but is barred from taking that action by external constraints—among obstetrician-gynecologists (OB-GYNs) in states with abortion bans. We conducted qualitative, semi-structured interviews with 54 OB-GYNs practicing in 13 of 14 states where abortion was illegal as of March 2023. Using a qualitative descriptive coding approach, we identified four types of clinical situations leading to moral distress: delaying treatment for patients with obstetric complications, conflict with other clinicians, denying care they would have provided locally prior to Dobbs, and restrictions on clinical counseling. These situations provoked feelings of anger, frustration, helplessness, and emotional exhaustion. Participants attributed moral distress to the cumulative toll of routinely being unable to provide evidence-based healthcare, in addition to the acute burden of managing obstetric emergencies in legal gray areas, which was relatively rarer. The findings demonstrate a previously unreported source of moral distress: the everyday chipping away of professional integrity that occurs when OB-GYNs cannot care for patients in the way that patients need. Next steps include developing institutional-level policies and programs to support clinicians and enable them to practice ethical medicine in abortion-restrictive policy environments.

The Investigation of a Nelsonian Approach to Socratic Dialogue with Student-Teachers at a Midwestern Private University

Student-teachers must be afforded opportunities to seek guidance on challenges that arise during teaching experiences, such as addressing student behavior, creating meaningful lessons, and fostering open and welcoming learning environments. Student-teachers can undergo negative experiences, which can lead to frustration, anger, and lack of interest in the profession. One way to address these challenges is through dialogue (Kumar & Downey, 2018). In this qualitative study, the researcher investigated whether a nuanced version of the Socratic method, developed by Leonard Nelson (1882-1927), can be an efficacious instructional strategy for finding solutions to student-teachers’ classroom challenges. The research study consisted of five undergraduate students enrolled in a student teaching seminar at a private midwestern university. The course prepared students for roles as new teachers by reading, discussing, and sharing challenges about teaching. Both the researcher and research participants worked together to faithfully follow the essential steps of a Nelsonian Socratic method approach. Once the preparation was settled, the researcher, as the external analyst, facilitated a Nelsonian Socratic dialogue with the student-teachers on a specific challenge a student teacher within the group was facing. After the dialogue, participants were assembled in a focus group and asked several questions about their experiences participating in a Nelsonian Socratic dialogue. Descriptive, categorical, and analytical coding of data indicated that participants reached consensus on a solution for the student-teacher’s challenge and perceived the Nelsonian Socratic method approach as an effective dialogic strategy that generated feelings of belongingness and empowerment as a group, a shared understanding about the challenge, and a healthy respect for dialogue.

Interests, Plans, and Hopes for Life After High School From Autistic Young Adults’ Perspectives

Obtaining the perspectives of autistic young adults is critical for planning educational and service delivery systems focused on the preparation for a high life quality after high school. The perspective and voice of autistic young adults who participated in the Center on Secondary Education for Students with Autism (CSESA) multi-site follow-up study conducted across three states (North Carolina, Wisconsin, and California) are described. Participants who exited high school or were enrolled in a district transition program responded to short-answer items on a Young Adult Questionnaire and Interview ( n = 148) and completed the Adolescent and Young Adult Activity Card Sort (Berg, 2015; n = 150) to address research questions about their current interests, future plans, and perceived barriers to fulfilling future plans using descriptive statistics and coding of interview responses. The young adults report current engagement in leisure activities, and 79% seek a job change and the opportunity to date and engage in a long-term relationship.

The experiences of medical and pharmacy students participating in a collaborative online international learning on social accountability and global health: a qualitative study

BackgroundSocial accountability aims to promote a collective ethic that upholds the fundamental values of equity, efficiency, solidarity, and social justice in healthcare and is now considered as a critical mission of academic health centers. Collaborative Online International Learning (COIL) is a pedagogical approach that uses digital technology to provide experiential international learning, specifically for increasingly diverse and multicultural healthcare work environments. The SOLID’AIRS program is an innovative French-language COIL that aims to set up international exchanges and workshops on social accountability between health sciences students. The aim of our study was to assess the impact of participation in the SOLID’AIRS program on medical and pharmacy students.MethodFive universities in four different countries were involved in SOLID’AIRS. We conducted a qualitative study by performing individual, semi-directed interviews with students who participated in the program from 2021 to 2022. A thematic analysis was conducted in five chronological phases: (1) reading, (2) descriptive coding, (3) conceptual coding, (4) identification of themes and (5) production of a coherent thematic structure.ResultsAfter including sixteen student participants, 13 medical and 3 pharmacy students, we identified four main themes related to the impact of participation: (1) previous experiences in social accountability and international learning, (2) perception of the program, (3) perceived impacts of the program, and (4) difficulties encountered and avenues to improve the program. Overall, the program was well received by all participants who reported the advantages and limitations of the online learning format. The primary advantage of this format was its feasibility. The participants noted both professional and personal benefits of the program for their current and future practice, including greater reflexivity towards health sciences practice. Based on the challenges faced during the program, particularly in coordinating group work and communication, participants suggested increased supervision of group projects by collaborators, and organizing at least one in-person meeting for future editions. The participants reported encountering difficulties during the COIL and suggested ways of improvement.ConclusionParticipating in a COIL on social accountability appears to be an effective way to adopt a reflective approach to medical practice and should be implemented and evaluated in other educational contexts.

Students' Learning Interest In Following The Process Pjok Learning At Mi Miftahul Ulum Plinggisan

This research uses quantitative descriptive methods to measure and analyxe The level of student interest in learning following the PJOK learning process at MI Miftahul Ulum Plinggisan. This method aims to get a clear picture of the factors that influence students' learning interest and the effectiveness of the strategies that have been implemented. The population of this study was all students at MI Miftahul Ulum Plinggisan, with samples selected randomly to ensure representation from each class. The research instrument is a questionnaire consisting of four main parts: Learning Motivation, Enjoyment, Involvement, and Perceptions of PJOK Lessons. Each section consists of several statements which are measured using a Likert scale with a range of 1 (Strongly Disagree) to 5 (Strongly Agree). Data collection was carried out by distributing questionnaires to students who were research samples, with filling in under the supervision of researchers to ensure completeness and accuracy of the data. The collected data was analyzed using descriptive statistical techniques, including editing and coding, tabulation, calculating descriptive statistics (mean, median, mode, and standard deviation), as well as Pearson correlation analysis to see the relationship between the variables motivation, enjoyment, involvement, and perception with interest. student learning. The research results showed that as many as 70% of students felt motivated to take part in PJOK lessons, as many as 80% of students enjoyed the activities in learning PJOK, as many as 80% of students were active in PJOK activities. As many as 75% of students have a positive view of the importance of PJOK lessons.

Ensemble neural models for ICD code prediction using unstructured and structured healthcare data

Disease coding is the process of assigning one or more standardized diagnostic codes to clinical notes that are maintained by health practitioners (e.g. clinicians) to track patient condition. Such a coding process is often expensive and error-prone, as human medical coders primarily perform it. Automating diagnostic coding using Artificial Intelligence is seen as an essential solution in Hospital Information Management Systems and approaches built on Convolutional Neural Networks currently perform best. In this work, a neural model built on unstructured clinical text for enabling automatic diagnostic coding for given patient discharge summaries is proposed. We incorporate a structured self-attention mechanism designed to boost learning of label-specific vectors and the significant clinical text snippets associated with a certain label for this purpose. These vectors are then combined with a novel code description pipeline leveraging the descriptions provided for each standardized diagnostic code. The proposed model achieved best performance in terms of standard metrics over the MIMIC-III dataset, outperforming models based on Longformers and Knowledge graphs. Furthermore, to leverage structured clinical data to enhance the proposed model, and to enable improved diagnostic code prediction, model ensembling is considered. A neural model built on structured data by leveraging supervised machine learning algorithms such as random forest and boosting, is designed for multi-class code classification. Experimental results revealed that the proposed ensemble models show promising performance compared to traditional models that rely solely on unstructured or structured clinical data, emphasizing their suitability for real-world deployment.

Interactions with new-to-teacher language resources: Supporting translingual composing in a multilingual elementary classroom

Educators should support multilingual students’ translingual writing. However, it can be challenging for teachers to support students’ composing in languages that teachers do not speak. Drawing on a community translanguaging lens, this paper explores this issue by asking: How did teachers talk about and interact with language resources that were new to them while supporting translingual writing in an English-medium classroom? Data were collected using ethnographic and practitioner research methods across 1 year in one second-grade writing workshop in the U.S. Students spoke Spanish, Korean, French, Tagalog, or English, and classroom teachers spoke English and Spanish. Data analysis first involved descriptive coding of videorecorded composing interactions to identify every teacher interaction involving a new-to-them named language. These events were then re-examined using constant comparative coding to identify interactional patterns. This yielded three main findings; teachers: (1) positioned students as language experts and themselves as language learners, (2) drew on shared language resources to support student writing in new-to-teacher languages, and (3) expanded audiences to support student writing in new-to-teacher languages. Implications include pedagogical steps teachers can take to support students’ use of new-to-teacher languages while writing, and ideological and social implications of teacher talk about those languages.

Sharing practices of software artefacts and source code for reproducible research

AbstractWhile source code of software and algorithms depicts an essential component in all fields of modern research involving data analysis and processing steps, it is uncommonly shared upon publication of results throughout disciplines. Simple guidelines to generate reproducible source code have been published. Still, code optimization supporting its repurposing to different settings is often neglected and even less thought of to be registered in catalogues for a public reuse. Though all research output should be reasonably curated in terms of reproducibility, it has been shown that researchers are frequently non-compliant with availability statements in their publications. These do not even include the use of persistent unique identifiers that would allow referencing archives of code artefacts at certain versions and time for long-lasting links to research articles. In this work, we provide an analysis on current practices of authors in open scientific journals in regard to code availability indications, FAIR principles applied to code and algorithms. We present common repositories of choice among authors. Results further show disciplinary differences of code availability in scholarly publications over the past years. We advocate proper description, archiving and referencing of source code and methods as part of the scientific knowledge, also appealing to editorial boards and reviewers for supervision.

Leadership for action on climate change: the role of education policy and professional standards in the Hong Kong context

PurposeUsing Hong Kong as the context of study, this paper examines the role of education policy, particularly professional standards for teachers (PST) and principals (PSP), in enabling leadership for action on contemporary challenges and crises such as climate change.Design/methodology/approachThis paper utilises critical policy analysis (CPA) and an associated analytical framework to examine key policy texts that are designed with the intent to frame the professional roles, practice and learning of educators and leaders in Hong Kong. The data analysis process involved rounds of descriptive coding, which was then collated into broader themes for discussion.FindingsPolicy texts emphasised the scope for educators to exercise leadership and decision-making that supports innovation, improvement and collaboration in the pursuit of a broad social mission for education, emphasising the skills students do and will need to respond to dynamic contemporary challenges. However, how this relates to broader policy goals, and limited articulation of the means through which this is possible, questions remain about the extent to which teachers and leaders are able to exercise leadership for action on climate change.Originality/valueThis study has important implications for how those in policymaking, research and practice spaces understand the framing of the role of teachers and school leaders in policy and the influence of this on their capacity to lead action on contemporary challenges and crises, such as climate change.

Coding Methods for Enhanced Manual Data Transcripts: Contributing to Sustainable Development Goals (SDG)

Objective: The objective of this study is to explore the graduate school writers' intentions and behaviors related to adopting different manual coding methods. Theoretical Framework: In this topic, the main concepts and theories that underpin the research are presented. In this study, the Theory of Planned Behavior (TPB )stand out, providing a solid basis for understanding the context of the investigation. Method: The methodology adopted for this research comprises Multi-phase Mixed-method; Frequency, Mean and Standard Deviation for quantitative was employed; Different Coding such as Emotion, In-vivo, Descriptive and Axial Coding Methods were also utilized for qualitative part. A probability sampling approach was employed. The Coding Manual for qualitative researchers by Saldana (2014) was adopted for instrumentation. Results and Discussion: The survey reveals a moderate level of understanding and proficiency in manual coding methods and theming. While respondents excel in foundational coding principles, there is room for improvement in using multiple coding methods and in vivo coding. The findings suggest that writers predominantly rely on traditional coding methods, revealing a gap in understanding and application of effective coding techniques and thematic analysis. Bridging the gap between traditional and alternative coding methods is crucial to advancing research practices in graduate school settin. Research Implications: The practical and theoretical implications of this research are discussed, providing insights into how the results can be applied or influence practices in the field of Research. By addressing mentioned challenges, writers can improve the accuracy and depth of their data analysis, leading to more robust and meaningful research outcomes. In conclusion, there is a need to bridge the gap between the traditional coding method currently employed by graduate school writers and the alternative coding methods available. Additionally, providing training and support to enhance their understanding and application of coding methods and thematic analysis will contribute to the advancement of their research practices and ultimately lead to more comprehensive and insightful findings Originality/Value: This study contributes to the literature by providing valuable insights into the prevailing reliance on traditional coding methods among graduate school writers, thereby addressing Sustainable Development Goal (SDG) 4 on quality education. It highlights the associated challenges in effectively coding and theming qualitative data, revealing a gap between awareness and application of alternative coding techniques. The research highlight the urgent need for enhanced training and support in research practices, advocating for increased adoption of more effective coding methods. By improving the accuracy and depth of data analysis, this initiative not only aims to address current shortcomings but also seeks to advance qualitative research practices, fostering more robust and insightful research outcomes in academia

Injectable Contraceptives as an Underutilized Option for Women's Reproductive Health: An Exploratory Qualitative Study.

Injectable contraceptives (IC) provide a highly effective, reversible method of preventing conception, yet discontinuation rates are high. Health workers play a crucial role in the successful implementation of family welfare services. Adding up the basket of choices without knowing the community's needs can lead to poor utilization of services. To explore the facilitators and barriers to the utility of injectable contraceptives among reproductive women from the user's point of view and to understand solutions from the key informants. The study was conducted in the field practice areas among reproductive women attending a primary health center in Puducherry. It was an exploratory qualitative study in which in-depth interviews were conducted among 19 IC users using an interview guide. This was followed by a key informant interview with seven service providers, including doctors, staff nurses, auxiliary nurse midwives, and Anganwadi workers, to explore the solutions for the identified barriers. Purposive, convenient sampling was adopted for the selection of study participants, and the sample size was chosen until the point of saturation. Two investigators trained in qualitative research have performed a manual content analysis of transcripts to ensure credibility. Descriptive codes were derived, and similar codes were merged into categories and themes. The most common facilitators were awareness from service providers and dissatisfaction with previous methods. Fear of side effects, inadequate information, out-of-pocket expenditure, lack of family support, and sociocultural myths were the most common barriers. Key informants suggested counseling on side effects, incentive-based follow-up, universal health insurance, couple-based counseling, training of service providers, and a positive deviance approach. Health workers are pivotal in the successful delivery of family welfare services. The acceptability of IC could be improved by addressing concerns about side effects and its effective management through various targeted interventions.

(Re)constructing and (re)negotiating identities: A case study of postsecondary language teachers and social justice education

AbstractThis study explores the identity formation of two postsecondary language teachers and course coordinators—one in German and one in French—as they created intermediate‐level, content‐based social justice curricular units for a multiyear project. Using a multiple‐case‐study methodology and cultural–historical activity theory, this qualitative investigation answers the following research question: How do experienced postsecondary language teachers (re)construct and (re)negotiate their identities while developing social justice teaching materials? Findings, based on multicycle descriptive coding, reveal that participants’ past experiences, personal values, feelings of self‐doubt, and positionality as content‐based language teachers influenced their language teacher identity, as did tensions within and across three activity systems: the curriculum development project, classroom teaching, and the coordination of multisection courses. These findings have important implications for teacher professional development and community building related to social justice in language education.

Semantic similarity loss for neural source code summarization

AbstractThis paper presents a procedure for and evaluation of using a semantic similarity metric as a loss function for neural source code summarization. Code summarization is the task of writing natural language descriptions of source code. Neural code summarization refers to automated techniques for generating these descriptions using neural networks. Almost all current approaches involve neural networks as either standalone models or as part of a pretrained large language models, for example, GPT, Codex, and LLaMA. Yet almost all also use a categorical cross‐entropy (CCE) loss function for network optimization. Two problems with CCE are that (1) it computes loss over each word prediction one‐at‐a‐time, rather than evaluating a whole sentence, and (2) it requires a perfect prediction, leaving no room for partial credit for synonyms. In this paper, we extend our previous work on semantic similarity metrics to show a procedure for using semantic similarity as a loss function to alleviate this problem, and we evaluate this procedure in several settings in both metrics‐driven and human studies. In essence, we propose to use a semantic similarity metric to calculate loss over the whole output sentence prediction per training batch, rather than just loss for each word. We also propose to combine our loss with CCE for each word, which streamlines the training process compared to baselines. We evaluate our approach over several baselines and report improvement in the vast majority of conditions.