Search results for: representations

Authors: Riikka Elina Kallio

Abstract:

“The walls have ears” phrase is defining the era of dictatorship in Portugal (1926–1974) and political unrest decades in Finland (1917–1948). The phrase is referring to the policing of the political, secret police, PIDE (Policia Internacional e de Defesa do Estado 1933–1969) in Portugal and VALPO (Valtiollinen Poliisi 1939–1948) in Finland. Free speech at any public space and even in private events could be fatal. The members of the PIDE/VALPO or informers/collaborators could be listening. Strict censorship under the Salazar´s regime was controlling media for example newspapers, music, and the film industry. Similarly, the politically affected censorship influenced the media in Finland in those unrest decades. This article examines the similarities and the differences in the images of the political police in Finland and Portugal, by analyzing Finnish and Portuguese films from the nineteen-thirties to nineteensixties. The text addresses two main research questions: what are the common and different features in the representations of the Finnish and Portuguese political police in films between the 1930s and 1960s, and how did the national censorship affect these representations? This study approach is interdisciplinary, and it combines film studies and criminology. Close reading is a practical qualitative method for analyzing films and in this study, close reading emphasizes the features of the police officer. Criminology provides the methodological tools for analysis of the police universal features and European common policies. The characterization of the police in this study is based on Robert Reiner´s 1980s and Timo Korander´s 2010s definitions of the police officer. The research material consisted of the Portuguese films from online film archives and Finnish films from Movie Making Finland -project´s metadata which offered suitable material by data mining the keywords such as poliisi, poliisipäällikkö and konstaapeli (police, police chief, police constable). The findings of this study suggest that even though there are common features of the images of the political police in Finland and Portugal, there are still national and cultural differences in the representations of the political police and policing.

Keywords: censorship, film studies, images, PIDE, political police, VALPO

Procedia PDF Downloads 29

Authors: Joseph Gear, Yue Xu, Ernest Foo, Praveen Gauravaran, Zahra Jadidi, Leonie Simpson

Abstract:

Deep learning methods have been seeing an increasing application to the long-standing security research goal of automatic vulnerability detection for source code. Attention, however, must still be paid to the task of producing vector representations for source code (code embeddings) as input for these deep learning models. Graphical representations of code, most predominantly Abstract Syntax Trees and Code Property Graphs, have received some use in this task of late; however, for very large graphs representing very large code snip- pets, learning becomes prohibitively computationally expensive. This expense may be reduced by intelligently pruning this input to only vulnerability-relevant information; however, little research in this area has been performed. Additionally, most existing work comprehends code based solely on the structure of the graph at the expense of the information contained by the node in the graph. This paper proposes Semantic-enhanced Code Embedding for Vulnerability Discovery (SCEVD), a deep learning model which uses semantic-based feature selection for its vulnerability classification model. It uses information from the nodes as well as the structure of the code graph in order to select features which are most indicative of the presence or absence of vulnerabilities. This model is implemented and experimentally tested using the SARD Juliet vulnerability test suite to determine its efficacy. It is able to improve on existing code graph feature selection methods, as demonstrated by its improved ability to discover vulnerabilities.

Keywords: code representation, deep learning, source code semantics, vulnerability discovery

Procedia PDF Downloads 122

Authors: Pragya Paneru

Abstract:

This paper explores ten primary teachers’ views on gender representation in primary-level textbooks altogether. Data was collected from the teachers who taught in private schools in Kailali and Kathmandu District. This research uses a semi-structured interview method to obtain information regarding teachers’ attitudes toward gender representations in textbook content. The interview data were analysed by using critical skills of qualitative research analysis methods, as suggested by Saldana and Omasta (2018). The findings revealed that most of the teachers were unaware and regarded gender issues as insignificant to discuss in primary-level classes. Most of them responded to the questions personally and claimed that there were no gender issues in their classrooms. Some of the teachers connected gender issues with contexts other than textbook representations, such as school discrimination in the distribution of salary among male and female teachers, school practices of awarding girls rather than boys as the most disciplined students, following girls’ first rule in the assembly marching, encouraging only girls in the stage shows, and involving students in gender-specific activities such as decorating works for girls and physical tasks for boys. The interview also revealed teachers’ covert gendered attitudes in their remarks. Nevertheless, most of the teachers accepted that gender-biased contents have an impact on learners, and this problem can be solved with more gender-centred research in the education field, discussions, and training to increase awareness regarding gender issues. Agreeing with the suggestion of teachers, this paper recommends proper training and awareness regarding how to confront gender issues in textbooks.

Keywords: content analysis, gender equality, school education, critical awareness

Procedia PDF Downloads 63

Authors: Alejandra G. Barbón, José Vila, Dania Vazquez

Abstract:

This study seeks to contribute to the advancement of library and archival sciences in the areas of records management, knowledge organization, and information architecture, particularly focusing on the enhancement of bibliographical description through the incorporation of visual interactive designs aimed to enrich the library users’ experience. In an era of heightened awareness about the legacy of hiddenness across special and rare collections in libraries and archives, along with the need for inclusivity in academia, the University of Miami Libraries has embarked on an innovative project that intersects the realms of print culture, decolonization, and digital technology. This proposal presents an exciting initiative to revitalize the study of Artists’ Books collections by employing digital visual representations to decolonize bibliographic records of some of the most unique materials and foster a more holistic understanding of cultural heritage. Artists' Books, a dynamic and interdisciplinary art form, challenge conventional bibliographic classification systems, making them ripe for the exploration of alternative approaches. This project involves the creation of a digital platform that combines multimedia elements for digital representations, interactive information retrieval systems, innovative information architecture, trending bibliographic cataloging and metadata initiatives, and collaborative curation to transform how we engage with and understand these collections. By embracing the potential of technology, we aim to transcend traditional constraints and address the historical biases that have influenced bibliographic practices. In essence, this study showcases a groundbreaking endeavor at the University of Miami Libraries that seeks to not only enhance bibliographic practices but also confront the legacy of hiddenness across special and rare collections in libraries and archives while strengthening conventional bibliographic description. By embracing digital visualizations, we aim to provide new pathways for understanding Artists' Books collections in a manner that is more inclusive, dynamic, and forward-looking. This project exemplifies the University’s dedication to fostering critical engagement, embracing technological innovation, and promoting diverse and equitable classifications and representations of cultural heritage.

Keywords: decolonizing bibliographic cataloging frameworks, digital visualizations information architecture platforms, collaborative curation and inclusivity for records management, engagement and accessibility increasing interaction design and user experience

Procedia PDF Downloads 43

Authors: Despoina Mouratidis, Katia Lida Kermanidis

Abstract:

Machine translation (MT) is a very important task in Natural Language Processing (NLP). MT evaluation is crucial in MT development, as it constitutes the means to assess the success of an MT system, and also helps improve its performance. Several methods have been proposed for the evaluation of (MT) systems. Some of the most popular ones in automatic MT evaluation are score-based, such as the BLEU score, and others are based on lexical similarity or syntactic similarity between the MT outputs and the reference involving higher-level information like part of speech tagging (POS). This paper presents a language-independent machine learning framework for classifying pairwise translations. This framework uses vector representations of two machine-produced translations, one from a statistical machine translation model (SMT) and one from a neural machine translation model (NMT). The vector representations consist of automatically extracted word embeddings and string-like language-independent features. These vector representations used as an input to a multi-layer neural network (NN) that models the similarity between each MT output and the reference, as well as between the two MT outputs. To evaluate the proposed approach, a professional translation and a "ground-truth" annotation are used. The parallel corpora used are English-Greek (EN-GR) and English-Italian (EN-IT), in the educational domain and of informal genres (video lecture subtitles, course forum text, etc.) that are difficult to be reliably translated. They have tested three basic deep learning (DL) architectures to this schema: (i) fully-connected dense, (ii) Convolutional Neural Network (CNN), and (iii) Long Short-Term Memory (LSTM). Experiments show that all tested architectures achieved better results when compared against those of some of the well-known basic approaches, such as Random Forest (RF) and Support Vector Machine (SVM). Better accuracy results are obtained when LSTM layers are used in our schema. In terms of a balance between the results, better accuracy results are obtained when dense layers are used. The reason for this is that the model correctly classifies more sentences of the minority class (SMT). For a more integrated analysis of the accuracy results, a qualitative linguistic analysis is carried out. In this context, problems have been identified about some figures of speech, as the metaphors, or about certain linguistic phenomena, such as per etymology: paronyms. It is quite interesting to find out why all the classifiers led to worse accuracy results in Italian as compared to Greek, taking into account that the linguistic features employed are language independent.

Keywords: machine learning, machine translation evaluation, neural network architecture, pairwise classification

Procedia PDF Downloads 100

Authors: Eirini Arvanitaki

Abstract:

Romance has incrementally grown in popularity over the last century. The first Mills & Boon romance novel was published in 1909 and since then romance has not only survived but it has become a long standing genre. There are several reasons behind its durability and success. First, its ability to please and appeal to a mass audience. Romance novels are products of commercial success situated in large scale production, especially if one takes into account the high volume of romance novels published, translated and distributed all around the world every month. Second, what has also contributed to keeping the romance genre alive is the content of the books and their effect on the reader. These are stories of two heterosexual individuals who meet, fall in love, face obstacles and successfully overcome them. Through the love plots, the books address anxieties, concerns and everyday troubles that the average reader can identify with. Additionally, the romance novel is a means of escapism from everyday life and responsibilities as well as a short-lived opportunity to enjoy personal time/space and focus on one’s self. Third, the genre’s ability to adapt to the periods and societies in which it is published has also assisted in prolonging its longevity. This paper discusses the ways in which popular romance authors write and engage with the body. Despite the claim that popular romance narratives adjust their contents in accordance with different time periods and social phenomena, the paper highlights the dissimilarities between writing the female and male body and suggests that women romance writers are yet to break free from phallogocentric law. The examination of the projections of the body and the language used to describe it indicates that these narratives are flexible enough to adjust to twenty-first century but only within the limits of their own conventionality.

Keywords: body, gender, language, literary representations, popular romance narratives, taboo

Procedia PDF Downloads 77

Authors: Yuanchao Sang, Zhihao Gong, Longsheng Chen, Long Chen

Abstract:

Clothing fashion represents a human’s aesthetic appreciation towards everyday outfits and appetite for fashion, and it reflects the development of status in society, humanity, and economics. However, modelling fashion by machine is extremely challenging because fashion is too abstract to be efficiently described by machines. Even human beings can hardly reach a consensus about fashion. In this paper, we are dedicated to answering a fundamental fashion-related problem: what image feature best describes clothing fashion? To address this issue, we have designed and evaluated various image features, ranging from traditional low-level hand-crafted features to mid-level style awareness features to various current popular deep neural network-based features, which have shown state-of-the-art performance in various vision tasks. In summary, we tested the following 9 feature representations: color, texture, shape, style, convolutional neural networks (CNNs), CNNs with distance metric learning (CNNs&DML), AutoEncoder, CNNs with multiple layer combination (CNNs&MLC) and CNNs with dynamic feature clustering (CNNs&DFC). Finally, we validated the performance of these features on two publicly available datasets. Quantitative and qualitative experimental results on both intra-domain and inter-domain fashion clothing image retrieval showed that deep learning based feature representations far outweigh traditional hand-crafted feature representation. Additionally, among all deep learning based methods, CNNs with explicit feature clustering performs best, which shows feature clustering is essential for discriminative fashion feature representation.

Keywords: convolutional neural network, feature representation, image processing, machine modelling

Procedia PDF Downloads 107

Authors: Pablo De Macedo Silveira Vallejos

Abstract:

This article aims to observe autobiographical and visual narrative practices among users on Instagram. In this way, the work proposes to reflect on how image resources are used to develop edited representations of the self in that social network. The research aims to explore the uses of editing and the degrees of fictionalization present on Instagram.

Keywords: autobiography, visual narratives, representation, fiction, social media

Procedia PDF Downloads 37

Authors: Pamela Aimee Rigor

Abstract:

This study aims to examine the representation of Syrian refugees resettled in Vancouver and the Lower Mainland in local community and major newspapers. While there is strong support for immigration in Canada, public opinion towards refugees and asylum seekers is a bit more varied. Concerns about the legitimacy of refugee claims are among the common concerns of Canadians, and hateful or negative narratives are still present in Canadian media discourse which affects how people view refugees. To counter the narratives, these Syrian refugees must publicly declare how grateful they are because they are resettled in Canada. The dominant media discourse is that these refugees should be grateful as they have been graciously accepted by Canada and Canadians, once again upholding the image of Canada being a generous and humanitarian nation. The study examined the representation of Syrian refugees and the Syrian refugee resettlement in Canadian newspapers from September 2015 to October 2017 – around the time Prime Minister Trudeau came into power up until the present. Using a combination of content and discourse analysis, it aimed to uncover how local community and major newspapers in Vancouver covered the Syrian refugee ‘crisis’ – more particularly, the arrival and resettlement of the refugees in the country. Using the qualitative data analysis software Nvivo 12, the newspapers were analyzed and sorted into themes. Based on the initial findings, the discourse of Canada being a humanitarian country and Canadians being generous, as well as the idea of Syrian refugees having to publicly announce how grateful they are, is still present in the local community newspapers. This seems to be done to counter the hateful narratives of citizens who might view them as people who are abusing help provided by the community or the services provided by the government. However, compared to the major and national newspapers in Canada, many these local community newspapers are very inclusive of Syrian refugee voices. Most of the News and Community articles interview Syrian refugees and ask them their personal stories of plight, survival, resettlement and starting a ‘new life’ in Canada. They are not seen as potential threats nor are they dismissed – the refugees were named and were allowed to share their personal experiences in these news articles. These community newspapers, even though their representations are far from perfect, actually address some aspects of the refugee resettlement issue and respond to their community’s needs. There are quite a number of news articles that announce community meetings and orientations about the Syrian refugee crisis, ways to help in the resettlement process, as well as community fundraising activities to help sponsor refugees or resettle newly arrived refugees. This study aims to promote awareness of how these individuals are socially constructed so we can, in turn, be aware of the certain biases and stereotypes present, and its implications on refugee laws and public response to the issue.

Keywords: forced migration and conflict, media representations, race and multiculturalism, refugee studies

Procedia PDF Downloads 208

Authors: Brian Kwok

Abstract:

Hong Kong is well-known for its name as "the Pearl of the Orient", due to its spectacular night-view with vast amount of decorative neon lights on the streets. Neon signs are first used as the pervasive media of communication for all kinds of commercial advertising, ranging from movie theatres to nightclubs and department stores, and later appropriated by artists as medium of artwork. As a well-established visual language, it displays texts in bilingual format due to British's colonial influence, which are sometimes arranged in an opposite reading order. Research on neon signs as a visual representation is rare but significant because they are part of people’s collective memories of the unique cityscapes which associate the shifting values of people's daily lives and culture identity. Nevertheless, with the current policy to remove abandoned neon signs, their total number dramatically declines recently. The Buildings Department found an estimation of 120,000 unauthorized signboards (including neon signs) in Hong Kong in 2013, and the removal of such is at a rate of estimated 1,600 per year since 2006. In other words, the vernacular cultural values and historical continuity of neon signs will gradually be vanished if no immediate action is taken in documenting them for the purpose of research and cultural preservation. Therefore, the Hong Kong Neon Signs Archive project was established in June of 2015, and over 100 neon signs are photo-documented so far. By content analysis, this project will explore the two components of neon signs – the use of visual languages and vernacular tacit knowledge of neon makers. It attempts to answer these questions about Hong Kong's neon signs: 'What are the ways in which visual representations are used to produce our cityscapes and streetscapes?'; 'What are the visual languages and conventions of usage in different business types?'; 'What the intact knowledge are applied when producing these visual forms of neon signs?'

Keywords: cityscapes, neon signs, tacit knowledge, visual representation

Procedia PDF Downloads 262

Authors: Dimitris Varsamis, Nicholas Karampetakis

Abstract:

It is well known that any interpolating polynomial P(x,y) on the vector space Pn,m of two-variable polynomials with degree less than n in terms of x and less than m in terms of y has various representations that depends on the basis of Pn,m that we select i.e. monomial, Newton and Lagrange basis etc. The aim of this paper is twofold: a) to present transformations between the coordinates of the polynomial P(x,y) in the aforementioned basis and b) to present transformations between these bases.

Keywords: bivariate interpolation polynomial, polynomial basis, transformations, interpolating polynomial

Procedia PDF Downloads 368

Authors: Zana Llonçari Osmani, Nita Llonçari

Abstract:

In Kosovo, public rhetoric and popular sentiment position the War of 1998-99 (the war) as central to the formation of contemporary Kosovo's national identity. This period was marked by the forced massive displacement of Kosovo Albanians, the destruction of entire settlements, the loss of family members, and the profound emotional trauma experienced by civilians, particularly those who actively participated in the war as members of the Kosovo Liberation Army (KLA). Amidst these profound experiences, the Prekaz Massacre (The Massacre) is widely regarded as the defining event that preceded the final struggles of 1999 and the long-awaited attainment of independence. This study aims to explore how different visitors perceive the dominant discourse at The Memorial, a site dedicated to commemorating the Prekaz Massacre, and to identify the factors that influence their perceptions. The research employs a comprehensive mixed-method approach, combining online surveys, critical discourse analysis of visitor impressions, and content analysis of media representations. The findings of the study highlight the significant role played by original material remains in shaping visitor perceptions of The Memorial in comparison to the curated symbols and figurative representations interspersed throughout the landscape. While the design elements and physical layout of the memorial undeniably hold significance in conveying the memoryscape, there are notable shortcomings in enhancing the overall visitor experience. Visitors are still primarily influenced by the tangible remnants of the war, suggesting that there is room for improvement in how design elements can more effectively contribute to the memorial's narrative and the collective memory of the Prekaz Massacre.

Keywords: critical discourse analysis, memorialisation, national discourse, public rhetoric, war tourism

Procedia PDF Downloads 36

Authors: Elisabeth Pedersen

Abstract:

Representations and interpretations of womanhood and female sexualities bring forth various questions regarding gender norms, and the implications of these norms, which are permeating and repetitive within various societies. Literature is one form of media which provides the space to represent and interpret women, their bodies, and sexualities, and also reveals the power of language as an affective and affected force. As literature allows an opportunity to explore history and the representations of gender, power dynamics, and sexuality through historical contexts, this paper uses engaged theory through a comparative analysis of two work of literature, The Duchess of Malfi by John Wester, and The Sound and the Fury by William Faulkner. These novels span across space and time, which lends to the theory that repetitive tropes of womanhood and female sexuality in literature are influenced by and have an influence on the hegemonic social order throughout history. It analyzes how the representation of the dichotomy of male chivalry and honor, and female purity are disputed and questioned when a woman is portrayed as sexually emancipated, and explores the historical context in which these works were written to examine how socioeconomic events challenged the hegemonic social order. The analysis looks at how stereotypical ideals of womanhood and manhood have damaging implications on women, as the structure of society provides more privilege and power to men than to women, thus creating a double standard for men and women in regards to sexuality, sexual expression, and rights to sexual desire. This comparative analysis reveals how strict gender norms are permeating and have negative consequences. However, re-reading stories through a critical lens can provide an opportunity to challenge the repetitive tropes of female sexuality, and thus lead to the embrace of the complexity of female sexuality and expression.

Keywords: femininity, literature, representation, sexuality

Procedia PDF Downloads 315

Authors: Caiwei Zhu, Remke Klapwijk

Abstract:

Spatial thinking, a way of thinking based on the understanding and reasoning of spatial concepts and representations, is embedded in science, technology, engineering, arts, and mathematics (STEAM) learning. Aside from many studies that successfully used targeted training to improve students’ spatial thinking skills, few have closely examined how spatial thinking can be trained in classroom settings. Design and technology education, which receives increasing attention towards its integration into formal curriculums, inherently encompasses a wide range of spatial activities, such as constructing mental representations of design ideas, mentally transforming objects and materials to form designs, visually communicating design plans through annotated drawings, and creating 2D and 3D design artifacts. Among different design topics, biomimicry offers a unique avenue for students to recognize and analyze the shapes and structures in nature. By mapping the forms of plants and animals onto functions, students gain inspiration to solve human design challenges. This study is one of the first to highlight opportunities for training spatial thinking in a biomimicry design project for primary school students. Embracing methodological principles of educational design-based research, this case study is conducted along with iterations in the design of the intervention and collaboration with teachers. Data are harvested from small groups of 10- to 12-year-olds at an international school in the Netherlands. Classroom videos, semi-structured interviews with students, design drawings and artifacts, formative assessment, and the pre- and post-intervention spatial test triangulate evidence for students' spatial thinking. In addition to contributing to a theory of integrating spatial thinking in the primary curriculum, mechanisms underlying such improvement in spatial thinking are explored and discussed.

Keywords: biomimicry, design and technology education, primary education, spatial thinking

Procedia PDF Downloads 33

Authors: Adrián P. Gómez Mañas

Abstract:

Autonomous vehicles have already transformed urban policies and cities. This is the main assumption of our research, which aims to understand how the representations of the possible arrival of autonomous vehicles already transform priorities or actions in transport and more largely, urban policies. This research is done within the framework of a Ph.D. doctorate directed by Professor Xavier Desjardins at the Sorbonne University of Paris. Our hypotheses are: (i) the perspectives, representations, and imaginaries on autonomous vehicles already affect the stakeholders of urban policies; (ii) the discourses on the opportunities or threats of autonomous vehicles reflect the current strategies of the stakeholders. Each stakeholder tries to integrate a discourse on autonomous vehicles that allows them to change as little as possible their current tactics and strategies. The objective is to eventually make a comparison between three different cases: Paris, United Arab Emirates, and Bogota. We chose those territories because their contexts are very different, but they all have important interests in mobility and innovation, and they all have started to reflect on the subject of self-driving mobility. The main methodology used is to interview actors of the metropolitan area (local officials, leading urban and transport planners, influent experts, and private companies). This work is supplemented with conferences, official documents, press articles, and websites. The objective is to understand: 1) What they know about autonomous vehicles and where does their knowledge come from; 2) What they expect from autonomous vehicles; 3) How their ideas about autonomous vehicles are transforming their action and strategy in managing daily mobility, investing in transport, designing public spaces and urban planning. We are going to present the research and some preliminary results; we will show that autonomous vehicles are often viewed by public authorities as a lever to reach something else. We will also present that speeches are very influenced by local context (political, geographical, economic, etc.), creating an interesting balance between global and local influences. We will analyze the differences and similarities between the three cases and will try to understand which are the causes.

Keywords: autonomous vehicles, self-driving mobility, urban planning, urban mobility, transport, public policies

Procedia PDF Downloads 144

Authors: Mathew Wakefield, Matthew Mitchell, Lisa Wise, Christopher McCarthy

Abstract:

The properties of memory representations in artificial neural networks have cognitive implications. Distributed representations that encode instances as a pattern of activity across layers of nodes afford memory compression and enforce the selection of a single point in instance space. These encoding schemes also appear to distort the representational space, as well as trading off the ability to validate that input information is within the bounds of past experience. In contrast, a localist representation which encodes some meaningful information into individual nodes in a network layer affords less memory compression while retaining the integrity of the representational space. This allows the validity of an input to be determined. The validity (or familiarity) of input along with the capacity of localist representation for multiple instance selections affords a memory sampling approach that dynamically balances the bias-variance trade-off. When the input is familiar, bias may be high by referring only to the most similar instances in memory. When the input is less familiar, variance can be increased by referring to more instances that capture a broader range of features. Using this approach in a localist instance memory network, an experiment demonstrates a relationship between representational conflict, generalization performance, and memorization demand. Relatively small sampling ranges produce the best performance on a classic machine learning dataset of visual objects. Combining memory validity with conflict detection produces a reliable confidence judgement that can separate responses with high and low error rates. Confidence can also be used to signal the need for supervisory input. Using this judgement, the need for supervised learning as well as memory encoding can be substantially reduced with only a trivial detriment to classification performance.

Keywords: artificial neural networks, representation, memory, conflict monitoring, confidence

Procedia PDF Downloads 96

Authors: Burak Oksuzoglu, Savas Yildirim, Ferhat Kutlu

Abstract:

Opinion mining (OM) is one of the natural language processing (NLP) problems to determine the polarity of opinions, mostly represented on a positive-neutral-negative axis. The data for OM is usually collected from various social media platforms. In an era where social media has considerable control over companies’ futures, it’s worth understanding social media and taking actions accordingly. OM comes to the fore here as the scale of the discussion about companies increases, and it becomes unfeasible to gauge opinion on individual levels. Thus, the companies opt to automize this process by applying machine learning (ML) approaches to their data. For the last two decades, OM or sentiment analysis (SA) has been mainly performed by applying ML classification algorithms such as support vector machines (SVM) and Naïve Bayes to a bag of n-gram representations of textual data. With the advent of deep learning and its apparent success in NLP, traditional methods have become obsolete. Transfer learning paradigm that has been commonly used in computer vision (CV) problems started to shape NLP approaches and language models (LM) lately. This gave a sudden rise to the usage of the pretrained language model (PTM), which contains language representations that are obtained by training it on the large datasets using self-supervised learning objectives. The PTMs are further fine-tuned by a specialized downstream task dataset to produce efficient models for various NLP tasks such as OM, NER (Named-Entity Recognition), Question Answering (QA), and so forth. In this study, the traditional and modern NLP approaches have been evaluated for OM by using a sizable corpus belonging to a large private company containing about 76,000 comments in Turkish: SVM with a bag of n-grams, and two chosen pre-trained models, multilingual universal sentence encoder (MUSE) and bidirectional encoder representations from transformers (BERT). The MUSE model is a multilingual model that supports 16 languages, including Turkish, and it is based on convolutional neural networks. The BERT is a monolingual model in our case and transformers-based neural networks. It uses a masked language model and next sentence prediction tasks that allow the bidirectional training of the transformers. During the training phase of the architecture, pre-processing operations such as morphological parsing, stemming, and spelling correction was not used since the experiments showed that their contribution to the model performance was found insignificant even though Turkish is a highly agglutinative and inflective language. The results show that usage of deep learning methods with pre-trained models and fine-tuning achieve about 11% improvement over SVM for OM. The BERT model achieved around 94% prediction accuracy while the MUSE model achieved around 88% and SVM did around 83%. The MUSE multilingual model shows better results than SVM, but it still performs worse than the monolingual BERT model.

Keywords: BERT, MUSE, opinion mining, pretrained language model, SVM, Turkish

Procedia PDF Downloads 107

Authors: L. Atif, C. Rosenthal-Sabroux, M. Grundstein

Abstract:

The requirements engineering process is a crucial phase in the design of complex systems. The purpose of our research is to present a collaborative problem-driven requirements engineering approach that aims at improving the design of a Decision Support System as an Analytics application. This approach has been adopted to design a Human Resource management DSS. The Requirements Engineering process is presented as a series of guidelines for activities that must be implemented to assure that the final product satisfies end-users requirements and takes into account the limitations identified. For this, we know that a well-posed statement of the problem is “a problem whose crucial character arises from collectively produced estimation and a formulation found to be acceptable by all the parties”. Moreover, we know that DSSs were developed to help decision-makers solve their unstructured problems. So, we thus base our research off of the assumption that developing DSS, particularly for helping poorly structured or unstructured decisions, cannot be done without considering end-user decision problems, how to represent them collectively, decisions content, their meaning, and the decision-making process; thus, arise the field issues in a multidisciplinary perspective. Our approach addresses a problem-driven and collaborative approach to designing DSS technologies: It will reflect common end-user problems in the upstream design phase and in the downstream phase these problems will determine the design choices and potential technical solution. We will thus rely on a categorization of HR’s problems for a development mirroring the Analytics solution. This brings out a new data-driven DSS typology: Descriptive Analytics, Explicative or Diagnostic Analytics, Predictive Analytics, Prescriptive Analytics. In our research, identifying the problem takes place with design of the solution, so, we would have to resort a significant transformations of representations associated with the HR Analytics application to build an increasingly detailed representation of the goal to be achieved. Here, the collective cognition is reflected in the establishment of transfer functions of representations during the whole of the design process.

Keywords: DSS, collaborative design, problem-driven requirements, analytics application, HR decision making

Procedia PDF Downloads 257

Authors: Puja Mahesh

Abstract:

Purpose: A good slogan captures the essence of your brand's promised consumer benefit in one short phrase. A good slogan conjures up positive imagery about your business or your product. A good slogan has the element of immediacy. Immediacy does not necessarily mean that the slogan will inspire consumers to run right out and buy your product. It does mean, however, that your slogan has an immediate cognitive impact. It forces your audience to "stop-and-think" after exposure as a necessary first step toward remembering your slogan promise. A good slogan is memorable and durability. When your slogan promise is occupying prime real estate in the consumer's subconscious, it aids in recall and activates preference for your brand when you want it -when consumers are ready to buy. The objective of current study is to understand the consumer perception of slogans/taglines of higher education sector in India. Design/Methodology/Approach: Survey of 500 consumers (largely comprising of youth) will be done using questionnaire. Universities and institutes will be chosen on the basis of various streams and Credible Rankings. The perception will be taken from the respondents on the basis of scale. Findings: Catchy phrases, rhymes, music, jingles, avatars (visual representations) and unique imagery are just a few of the mnemonic clutter-busting tactics commonly used in slogans to stand apart from the competition and to aid in memory recall. The study will reveal whether it is true that catchy phrases, rhymes, music, jingles, avatars (visual representations) and unique imagery across disciplines and universities help in building stronger brands. It will also be found whether consumers pay more attention to reputation of University/ College or brand identity. Originality/Value: Researcher has not come across any study of Consumer Perception of Slogans/Taglines of Higher Education Brands in India. Also, it would be interesting to understand Consumer Perception of various colleges/streams particularly Management colleges who invest a lot of time in branding exercise.

Keywords: consumer perception, higher education, slogans, taglines

Procedia PDF Downloads 377

Authors: Érica Lima

Abstract:

In the digital era, in which we have an avalanche of information, it is not new that the Internet has brought new modes of communication and knowledge access. Characterized by the multiplicity of discourses, opinions, beliefs and cultures, the web is a space of political-ideological dimensions where people (who often do not know each other) interact and create representations, deconstruct stereotypes, and redefine identities. Currently, the translator needs to be able to deal with digital spaces ranging from specific software to social media, which inevitably impact on his professional life. One of the most impactful ways of being seen in cyberspace is the participation in social networking groups. In addition to its ability to disseminate information among participants, social networking groups allow a significant personal and social exposure. Such exposure is due to the visibility of each participant achieved not only on its personal profile page, but also in each comment or post the person makes in the groups. The objective of this paper is to study the representations of translators and translation process on the Internet, more specifically in publications in two Brazilian groups of great influence on the Facebook: "Translators/Interpreters" and "Translators, Interpreters and Curious". These chosen groups represent the changes the network has brought to the profession, including the way translators are seen and see themselves. The analyzed posts allowed a reading of what common sense seems to think about the translator as opposed to what the translators seem to think about themselves as a professional class. The results of the analysis lead to the conclusion that these two positions are antagonistic and sometimes represent conflict of interests: on the one hand, the society in general consider the translator’s work something easy, therefore it is not necessary to be well remunerated; on the other hand, the translators who know how complex a translation process is and how much it takes to be a good professional. The results also reveal that social networking sites such as Facebook provide more visibility, but it takes a more active role from the translator to achieve a greater appreciation of the profession and more recognition of the role of the translator, especially in face of increasingly development of automatic translation programs.

Keywords: Facebook, social representation, translation, translator

Procedia PDF Downloads 117

Authors: Dylan Elliott, Katya Pertsova

Abstract:

Preposition errors are some of the most common errors created by L2 speakers. In addition, improving error correction and detection methods remains an open issue in the realm of Natural Language Processing (NLP). This research investigates whether the bidirectional encoder representations from the transformers model (BERT) have the potential to correct preposition errors accurately enough to be useful in error correction software. This research finds that BERT performs strongly when the scope of its error correction is limited to preposition choice. The researchers used an open-source BERT model and over three hundred thousand edited sentences from Wikipedia, tagged for part of speech, where only a preposition edit had occurred. To test BERT’s ability to detect errors, a technique known as multi-level masking was used to generate suggestions based on sentence context for every prepositional environment in the test data. These suggestions were compared with the original errors in the data and their known corrections to evaluate BERT’s performance. The suggestions were further analyzed to determine if BERT more often agreed with the judgements of the Wikipedia editors. Both the untrained and fined-tuned models were compared. Finetuning led to a greater rate of error-detection which significantly improved recall, but lowered precision due to an increase in false positives or falsely flagged errors. However, in most cases, these false positives were not errors in preposition usage but merely cases where more than one preposition was possible. Furthermore, when BERT correctly identified an error, the model largely agreed with the Wikipedia editors, suggesting that BERT’s ability to detect misused prepositions is better than previously believed. To evaluate to what extent BERT’s false positives were grammatical suggestions, we plan to do a further crowd-sourcing study to test the grammaticality of BERT’s suggested sentence corrections against native speakers’ judgments.

Keywords: BERT, grammatical error correction, preposition error detection, prepositions

Procedia PDF Downloads 110

Authors: Elisabetta Di Giovanni

Abstract:

This paper focuses on the contemporary phenomenon of Anti-Gypsyism which is widely diffused on social representations of the so called 'Gypsies'. In Europe and especially in Italy, media tend to reproduce racist stereotypes and prejudices through a xenophobic depiction of this ethnic group, often offering an ethnocentric point of view. From an anthropological perspective, Roma people are a minority group actually facing diasporic phenomena in all Europe, produced by the host society.

Keywords: Roma people, anti-gypsyism, ethnocentrism, mass media

Procedia PDF Downloads 380

Authors: Rimmy Yadav, Anmol Preet Kaur

Abstract:

Big Data has demonstrated the vast potential in streamlining, deciding, spotting business drifts in different fields, for example, producing, fund, Information Technology. This paper gives a multi-disciplinary diagram of the research issues in enormous information and its procedures, instruments, and system identified with the privacy, data storage management, network and energy utilization, adaptation to non-critical failure and information representations. Other than this, result difficulties and openings accessible in this Big Data platform have made.

Keywords: big data, privacy, data management, network and energy consumption

Procedia PDF Downloads 270

Authors: Massimo Miccoli, Luca Marangoni, Alberto Aniello Scaringi, Alessandro Marceddu, Alessandro Amicone

Abstract:

The susceptibility of deep neural networks (DNNs) to adversarial manipulations is a recognized challenge within the computer vision domain. Adversarial examples, crafted by adding subtle yet malicious alterations to benign images, exploit this vulnerability. Various defense strategies have been proposed to safeguard DNNs against such attacks, stemming from diverse research hypotheses. Building upon prior work, our approach involves the utilization of autoencoder models. Autoencoders, a type of neural network, are trained to learn representations of training data and reconstruct inputs from these representations, typically minimizing reconstruction errors like mean squared error (MSE). Our autoencoder was trained on a dataset of benign examples; learning features specific to them. Consequently, when presented with significantly perturbed adversarial examples, the autoencoder exhibited high reconstruction errors. The architecture of the autoencoder was tailored to the dimensions of the images under evaluation. We considered various image sizes, constructing models differently for 256x256 and 512x512 images. Moreover, the choice of the computer vision model is crucial, as most adversarial attacks are designed with specific AI structures in mind. To mitigate this, we proposed a method to replace image-specific dimensions with a structure independent of both dimensions and neural network models, thereby enhancing robustness. Our multi-modal autoencoder reconstructs the spectral representation of images across the red-green-blue (RGB) color channels. To validate our approach, we conducted experiments using diverse datasets and subjected them to adversarial attacks using models such as ResNet50 and ViT_L_16 from the torch vision library. The autoencoder extracted features used in a classification model, resulting in an MSE (RGB) of 0.014, a classification accuracy of 97.33%, and a precision of 99%.

Keywords: adversarial attacks, malicious images detector, binary classifier, multimodal transformer autoencoder

Procedia PDF Downloads 31

Authors: Ebipatei Victoria Tunyan, T. A. Cao, Cheol Young Ock

Abstract:

Detecting subjectively biased statements is a vital task. This is because this kind of bias, when present in the text or other forms of information dissemination media such as news, social media, scientific texts, and encyclopedias, can weaken trust in the information and stir conflicts amongst consumers. Subjective bias detection is also critical for many Natural Language Processing (NLP) tasks like sentiment analysis, opinion identification, and bias neutralization. Having a system that can adequately detect subjectivity in text will boost research in the above-mentioned areas significantly. It can also come in handy for platforms like Wikipedia, where the use of neutral language is of importance. The goal of this work is to identify the subjectively biased language in text on a sentence level. With machine learning, we can solve complex AI problems, making it a good fit for the problem of subjective bias detection. A key step in this approach is to train a classifier based on BERT (Bidirectional Encoder Representations from Transformers) as upstream model. BERT by itself can be used as a classifier; however, in this study, we use BERT as data preprocessor as well as an embedding generator for a Bi-LSTM (Bidirectional Long Short-Term Memory) network incorporated with attention mechanism. This approach produces a deeper and better classifier. We evaluate the effectiveness of our model using the Wiki Neutrality Corpus (WNC), which was compiled from Wikipedia edits that removed various biased instances from sentences as a benchmark dataset, with which we also compare our model to existing approaches. Experimental analysis indicates an improved performance, as our model achieved state-of-the-art accuracy in detecting subjective bias. This study focuses on the English language, but the model can be fine-tuned to accommodate other languages.

Keywords: subjective bias detection, machine learning, BERT–BiLSTM–Attention, text classification, natural language processing

Procedia PDF Downloads 91

Authors: R. Simões, M. Silveirinha

Abstract:

Driven by the need to understand the results of previous research that clearly shows deep unbalances of the media discourses about women and men in spite of the growing numbers of female journalists, our paper aims to progress from the 'what' to the 'why' of these unbalanced representations. Furthermore, it does so at a time when journalism is undergoing a dramatic change in terms of professional practices and in how media organizations are organized and run, affecting women in particular. While some feminist research points to the fact that female and male journalists evaluate the role of the news and production methods in similar ways feminist theorizing also suggests that thought and knowledge are highly influenced by social identity, which is also inherently affected by the experiences of gender. This is particularly important at a time of deep societal and professional changes. While there are persuasive discussions of gender identities at work in newsrooms in various countries studies on the issue will benefit from cases that focus on the particularities of local contexts. In our paper, we present one such case: the case of Portugal, a country hit hard by austerity measures that have affected all cultural industries including journalism organizations, already feeling the broader impacts of the larger societal changes of the media landscape. Can we gender these changes? How are they felt and understood by female and male journalists? And how are these discourses framed by androcentric, feminist and post-feminist sensibilities? Foregrounding questions of gender, our paper seeks to explore some of the interactions of societal and professional forces, identifying their gendered character and outlining how they shape journalism work in general and the production of unbalanced gender representations in particular. We do so grounded in feminist studies of journalism as well as feminist organizational and work studies, looking at a corpus of 20 in-depth interviews of female and male Portuguese journalists. The research findings illustrate how gender in journalism practices interacts with broader experiences of the cultural and economic contexts and show the ambivalences of these interactions in news organizations.

Keywords: gender, journalism, newsroom culture, Portuguese journalists

Procedia PDF Downloads 364

Authors: Sharmila Parmanand

Abstract:

This paper examines dominant media representations of prostitution in the Philippines and interrogates sex workers’ interactions with the media establishment. This analysis of how sex workers are constituted in media, often as both innocent victims and immoral actors, contributes to an understanding of public discourse on sex work in the Philippines, where decriminalisation has recently been proposed and sex workers are currently classified as potential victims under anti-trafficking laws but also as criminals under the penal code. The first part is an analysis of media coverage of two prominent themes on prostitution: first, raid and rescue operations conducted by law enforcement; and second, prostitution on military bases and tourism hotspots. As a result of pressure from activists and international donors, these two themes often define the policy conversations on sex work in the Philippines. The discourses in written and televised news reports and documentaries from established local and international media sources that address these themes are explored through content analysis. Conclusions are drawn based on specific terms commonly used to refer to sex workers, how sex workers are seen as performing their cultural roles as mothers and wives, how sex work is depicted, associations made between sex work and public health, representations of clients and managers and ‘rescuers’ such as the police, anti-trafficking organisations, and faith-based groups, and which actors are presumed to be issue experts. Images of how prostitution is used as a metaphor for relations between the Philippines and foreign nations are also deconstructed, along with common tropes about developing world female subjects. In general, sex workers are simultaneously portrayed as bad mothers who endanger their family’s morality but also as long-suffering victims who endure exploitation for the sake of their children. They are also depicted as unclean, drug-addicted threats to public health. Their managers and clients are portrayed as cold, abusive, and sometimes violent, and their rescuers as moral and altruistic agents who are essential for sex workers’ rehabilitation and restoration as virtuous citizens. The second part explores sex workers’ own perceptions of their interactions with media, through interviews with members of the Philippine Sex Workers Collective, a loose organisation of sex workers around the Philippines. They reveal that they are often excluded by media practitioners and that they do not feel that they have space for meaningful self-revelation about their work when they do engage with journalists, who seem to have an overt agenda of depicting them as either victims or women of loose morals. In their assessment, media narratives do not necessarily reflect their lived experiences, and in some cases, coverage of rescues and raid operations endangers their privacy and instrumentalises their suffering. Media representations of sex workers may produce subject positions such as ‘victims’ or ‘criminals’ and legitimize specific interventions while foreclosing other ways of thinking. Further, in light of media’s power to reflect and shape public consciousness, it is a valuable academic and political project to examine whether sex workers are able to assert agency in determining how they are represented.

Keywords: discourse analysis, news media, sex work, trafficking

Procedia PDF Downloads 347

Authors: Evangelia-Matroni Tomara

Abstract:

This thesis answers the question whether the media representations and reporting in 2015-2016 - especially, after the image of the drowned three-year-old Syrian boy in the Mediterranean Sea which made global headlines in the beginning of September 2015 -, the European Commission regulatory sources material and related reporting, have the power to challenge the conceptualization of humanitarianism or even redefine it. The theoretical foundations of the thesis are based on humanitarianism and its core definitions, the power of media representations and the relative portrayal of migrants, refugees and/or asylum seekers, as well as the dominant migration discourse and EU migration governance. Using content analysis for the media portrayal of migrants (436 newspaper articles) and qualitative content analysis for the European Commission Communication documents from May 2015 until June 2016 that required various depths of interpretation, this thesis allowed us to revise the concept of humanitarianism, realizing that the current crisis may seem to be a turning point for Europe but is not enough to overcome the past hostile media discourses and suppress the historical perspective of security and control-oriented EU migration policies. In particular, the crisis helped to shift the intensity of hostility and the persistence in the state-centric, border-oriented securitization in Europe into a narration of victimization rather than threat where mercy and charity dynamics are dominated and into operational mechanisms, noting the emergency of immediate management of the massive migrations flows, respectively. Although, the understanding of a rights-based response to the ongoing migration crisis, is being followed discursively in both political and media stage, the nexus described, points out that the binary between ‘us’ and ‘them’ still exists, with only difference that the ‘invaders’ are now ‘pathetic’ but still ‘invaders’. In this context, the migration crisis challenges the concept of humanitarianism because rights dignify migrants as individuals only in a discursive or secondary level while the humanitarian work is mostly related with the geopolitical and economic interests of the ‘savior’ states.

Keywords: European Union politics, humanitarianism, immigration, media representation, policy-making, refugees, security studies

Procedia PDF Downloads 251

Authors: Gulgassyl Nugmanova, Zhanat Zhunussova, Kuralay Yesmakhanova, Galya Mamyrbekova, Ratbay Myrzakulov

Abstract:

In this paper, we consider some integrable Heisenberg Ferromagnet Equations with self-consistent potentials. We study their Lax representations. In particular we derive their equivalent counterparts in the form of nonlinear Schr\"odinger type equations. We present the integrable reductions of the Heisenberg Ferromagnet Equations with self-consistent potentials. These integrable Heisenberg Ferromagnet Equations with self-consistent potentials describe nonlinear waves in ferromagnets with some additional physical fields.

Keywords: Heisenberg Ferromagnet equations, soliton equations, equivalence, Lax representation

Procedia PDF Downloads 422

Authors: Michael Lousis

Abstract:

Mathematical competence is the great aim of every mathematical teaching and learning endeavour. This can be defined as an idealised conceptualisation of the quality of cognition and the ability of implementation in practice of the mathematical subject matter, which is included in the curriculum, and is displayed only through performance of doing mathematics. The present study gives a clear definition of mathematical competence in the domains of Arithmetic and Algebra that stems from the explanation of the learners’ errors in these domains. The learners, whose errors are explained, were Greek and English participants of a large, international, longitudinal, comparative research program entitled the Kassel Project. The participants’ errors emerged as results of their work in dealing with mathematical questions and problems of the tests, which were presented to them. The construction of the tests was such as only the outcomes of the participants’ work was to be encompassed and not their course of thinking, which resulted in these outcomes. The intention was that the tests had to provide undeviating comparable results and simultaneously avoid any probable bias. Any bias could stem from obtaining results by involving so many markers from different countries and cultures, with so many different belief systems concerning the assessment of learners’ course of thinking. In this way the validity of the research was protected. This fact forced the implementation of specific research methods and theoretical prospects to take place in order the participants’ erroneous way of thinking to be disclosed. These were Methodological Pragmatism, Symbolic Interactionism, Philosophy of Mind and the ideas of Computationalism, which were used for deciding and establishing the grounds of the adequacy and legitimacy of the obtained kinds of knowledge through the explanations given by the error analysis. The employment of this methodology and of these theoretical prospects resulted in the definition of the learners’ mathematical competence, which is the thesis of the present study. Thus, learners’ mathematical competence is depending upon three key elements that should be developed in their minds: appropriate representations, appropriate meaning, and appropriate developed schemata. This definition then determined the development of appropriate teaching practices and interventions conducive to the achievement and finally the entailment of mathematical competence.

Keywords: representations, meaning, appropriate developed schemata, computationalism, error analysis, explanations for the probable causes of the errors, Kassel Project, mathematical competence

Procedia PDF Downloads 238