Search results for: 2-gram BLEU

Authors: Promise Ekpo Osaine, Daniel Melesse

Abstract:

Image captioning is an active area of research in the multi-modal artificial intelligence (AI) community as it connects vision and language understanding, especially in settings where it is required that a model understands the content shown in an image and generates semantically and grammatically correct descriptions. In this project, we followed a standard approach to a deep learning-based image captioning model, injecting architecture for the encoder-decoder setup, where the encoder extracts image features, and the decoder generates a sequence of words that represents the image content. As such, we investigated image encoders, which are ResNet101, InceptionResNetV2, EfficientNetB7, EfficientNetV2M, and CLIP. As a caption generation structure, we explored long short-term memory (LSTM). The CLIP-LSTM model demonstrated superior performance compared to the encoder-decoder models, achieving a BLEU-1 score of 0.904 and a BLEU-4 score of 0.640. Additionally, among the CNN-LSTM models, EfficientNetV2M-LSTM exhibited the highest performance with a BLEU-1 score of 0.896 and a BLEU-4 score of 0.586 while using a single-layer LSTM.

Keywords: multi-modal AI systems, image captioning, encoder, decoder, BLUE score

Procedia PDF Downloads 25

Authors: Sikiru Ademola Adewale, Joe Thomas, Bolanle Hafiz Matti, Tosin Ige

Abstract:

This project demonstrates the implementation and use of an encoder-decoder model to perform a many-to-many mapping of video data to text captions. The many-to-many mapping occurs via an input temporal sequence of video frames to an output sequence of words to form a caption sentence. Data preprocessing, model construction, and model training are discussed. Caption correctness is evaluated using 2-gram BLEU scores across the different splits of the dataset. Specific examples of output captions were shown to demonstrate model generality over the video temporal dimension. Predicted captions were shown to generalize over video action, even in instances where the video scene changed dramatically. Model architecture changes are discussed to improve sentence grammar and correctness.

Keywords: decoder, encoder, many-to-many mapping, video captioning, 2-gram BLEU

Procedia PDF Downloads 63

Authors: Sinarwati Mohamad Suhaili, Naomie Salim, Mohamad Nazim Jambli

Abstract:

Sequence-to-sequence (seq2seq) models augmented with attention mechanisms are playing an increasingly important role in automated customer service. These models, which are able to recognize complex relationships between input and output sequences, are crucial for optimizing chatbot responses. Central to these mechanisms are neural attention weights that determine the focus of the model during sequence generation. Despite their widespread use, there remains a gap in the comparative analysis of different attention weighting functions within seq2seq models, particularly in the domain of chatbots using the customer support Twitter (CST) dataset. This study addresses this gap by evaluating four distinct attention-scoring functions -dot, multiplicative/general, additive, and an extended multiplicative function with a tanh activation parameter- in neural generative seq2seq models. Utilizing the CST dataset, these models were trained and evaluated over 10 epochs with the AdamW optimizer. Evaluation criteria included validation loss and BLEU scores implemented under both greedy and beam search strategies with a beam size of k=3. Results indicate that the model with the tanh-augmented multiplicative function significantly outperforms its counterparts, achieving the lowest validation loss (1.136484) and the highest BLEU scores (0.438926 under greedy search, 0.443000 under beam search, k=3). These results emphasize the crucial influence of selecting an appropriate attention-scoring function in improving the performance of seq2seq models for chatbots. Particularly, the model that integrates tanh activation proves to be a promising approach to improve the quality of chatbots in the customer support context.

Keywords: attention weight, chatbot, encoder-decoder, neural generative attention, score function, sequence-to-sequence

Procedia PDF Downloads 45

Authors: Khadija Refouh

Abstract:

Culture bound-expressions have been a bottleneck for Natural Language Processing (NLP) and comprehension, especially in the case of machine translation (MT). In the last decade, the field of machine translation has greatly advanced. Neural machine translation NMT has recently achieved considerable development in the quality of translation that outperformed previous traditional translation systems in many language pairs. Neural machine translation NMT is an Artificial Intelligence AI and deep neural networks applied to language processing. Despite this development, there remain some serious challenges that face neural machine translation NMT when translating culture bounded-expressions, especially for low resources language pairs such as Arabic-English and Arabic-French, which is not the case with well-established language pairs such as English-French. Machine translation of opaque idioms from English into French are likely to be more accurate than translating them from English into Arabic. For example, Google Translate Application translated the sentence “What a bad weather! It runs cats and dogs.” to “يا له من طقس سيء! تمطر القطط والكلاب” into the target language Arabic which is an inaccurate literal translation. The translation of the same sentence into the target language French was “Quel mauvais temps! Il pleut des cordes.” where Google Translate Application used the accurate French corresponding idioms. This paper aims to perform NMT experiments towards better translation of opaque idioms using high quality clean multilingual corpus. This Corpus will be collected analytically from human generated idiom translation. AutoML translation, a Google Neural Machine Translation Platform, is used as a custom translation model to improve the translation of opaque idioms. The automatic evaluation of the custom model will be compared to the Google NMT using Bilingual Evaluation Understudy Score BLEU. BLEU is an algorithm for evaluating the quality of text which has been machine-translated from one natural language to another. Human evaluation is integrated to test the reliability of the Blue Score. The researcher will examine syntactical, lexical, and semantic features using Halliday's functional theory.

Keywords: multilingual corpora, natural language processing (NLP), neural machine translation (NMT), opaque idioms

Procedia PDF Downloads 104

Authors: Tahar Alimi, Rahma Boujelbane, Wiem Derouich, Lamia Hadrich Belguith

Abstract:

Machine translation task of low-resourced languages such as Arabic is a challenging task. Despite the appearance of sophisticated models based on the latest deep learning techniques, namely the transfer learning and transformers, all models prove incapable of carrying out an acceptable translation which includes Arabic dialects because they not official status. In this paper, a machine translation model designed to translate Arabic multidialectal content into Modern Standard Arabic (MSA), leveraging both new and existing parallel resources. The latter achieved the best results for both Levantine and Maghrebi dialects with BLEU score of 64.99.

Keywords: Arabic translation, dialect translation, fine-tune, msa translation, transformer, translation

Procedia PDF Downloads 8

Authors: Benyamin Ahmadnia, Javier Serrano

Abstract:

In this paper we compare two different approaches for translating from Persian to Spanish, as a language pair with scarce parallel corpus. The first approach involves direct transfer using an statistical machine translation system, which is available for this language pair. The second approach involves translation through English, as a pivot language, which has more translation resources and more advanced translation systems available. The results show that, it is possible to achieve better translation quality using English as a pivot language in either approach outperforms direct translation from Persian to Spanish. Our best result is the pivot system which scores higher than direct translation by (1.12) BLEU points.

Keywords: statistical machine translation, direct translation approach, pivot language translation approach, parallel corpus

Procedia PDF Downloads 458

Authors: Yufen Qin

Abstract:

Topic-to-Essay generation is a challenging task in Natural language processing, which aims to generate novel, diverse, and topic-related text based on user input. Previous research has overlooked the generation of articles under the constraints of event elements, resulting in issues such as incomplete event elements and logical inconsistencies in the generated results. To fill this gap, this paper proposes an event-constrained approach for a topic-to-essay generation that enforces the completeness of event elements during the generation process. Additionally, a language model is employed to verify the logical consistency of the generated results. Experimental results demonstrate that the proposed model achieves a better BLEU-2 score and performs better than the baseline in terms of subjective evaluation on a real dataset, indicating its capability to generate higher-quality topic-related text.

Keywords: event element, language model, natural language processing, topic-to-essay generation.

Procedia PDF Downloads 192

Authors: Abdallah Bouguettoucha, Derradji Chebli, Sara Aga, Agueniou Fazia

Abstract:

The objective of this study was to looking for alternative materials (low cost) for the adsorption of textile dyes and optimizes the type which gives optimum adsorption and provides an explanation of the mechanism involved in the adsorption process. Adsorption of Orange II and Methylene blue on H2SO4 traited cone of Pinus brutia, was carried out at different initial concentrations of the dye (20, 50 and 100 mg / L) and at tow initial pH, pH 1 and 10 respectively. The models of Langmuir, Freundlich and Sips were used in this study to analyze the obtained results of the adsorption isotherm. PCB-0M had high adsorption capacities namely 32.8967 mg/g and 128.1651 mg/g, respectively for orange II and methylene blue and further indicated that the removal of dyes increased with increase in the ionic strength of solution, this was attributed to aggregation of dyes in solution. The potential of H2SO4 traited cone of Pinus brutia, an easily available and low cost material, to be used as an alternative biosorbent material for the removal of a dyes, Orange II and Methylene Bleu, from aqueous solutions was therefore confirmed.

Keywords: Methylene blue, orange II, cones of pinus brutia, adsorption

Procedia PDF Downloads 254

Authors: M. Rilliard

Abstract:

Based on a Multiliteracies-inspired and sociolinguistically-informed advanced French composition class, this study employed autobiographical narratives from speakers traditionally considered non-legitimate models for L2 teaching purposes of inspiring students to develop an authentic L2 voice and to see themselves as legitimate L2 speakers. Students explored their L2 identities in French through a self-inspired fictional character. Two autobiographical narratives of identity quest by non-traditional French speakers provided them guidance through this process: the novel Le Bleu des Abeilles (2013) and the film Qu’Allah Bénisse la France (2014). Written and French oral productions for different genres, as well as metalinguistic reflections in English, were collected and analyzed. Results indicate that ideas and materials that were relatable to students, namely relatable experiences and relatable language, were most useful to them in developing their L2 voices and achieving authentic and legitimate L2 speakership. These results point towards the benefits of using non-traditional speakers as pedagogical models, as they serve to legitimize students’ sense of their own L2-speakership, which ultimately leads them towards a better, more informed, mastery of the language.

Keywords: foreign language classroom, L2 identity, L2 learning and teaching, L2 writing, sociolinguistics

Procedia PDF Downloads 100

Authors: Abdi Djamila, Haffar Hichem

Abstract:

The poly methylene bleu (PMB) has been successfully electro deposited on fluorine doped tin oxide (FTO) conducting glass as substrate. Its optical, electrical and photoelectrochemical characterizations have been carried out in order to show the performances of such polymer. The deposited film shows a good electric conductivity which is well confirmed by the low gap value determinated optically by UV–vis spectroscopy. Like all polymers the PMB presents an absorption difference in the visible range function of the polarization potential, it is expressed by the strong conjugation at oxidized state but is weakened with leucoform formation at reduced state. The electrochemical analysis of the films permit to show the cyclic voltamperogram with the anodic oxidation and cathodic reduction states of the polymer and to locate the corresponding energy levels HOMO and LUMO of this later. The electrochemical impedance spectroscopy permit to see the conductive character of such film and to calculate important parameters as Rtc and CPE. The study of the photoelectro activity of our polymer shows that under exposure to intermittent light source this later exhibit important photocurrents which enables it to be used in photo organic ells.

Keywords: polymethylene blue, electropolymerization, homo-lumo, photocurrents

Procedia PDF Downloads 246

Authors: Cheikh Bamba Dione, Alla Lo, Elhadji Mamadou Nguer, Siley O. Ba

Abstract:

In this paper, we propose two neural machine translation (NMT) systems (French-to-Wolof and Wolof-to-French) based on sequence-to-sequence with attention and transformer architectures. We trained our models on a parallel French-Wolof corpus of about 83k sentence pairs. Because of the low-resource setting, we experimented with advanced methods for handling data sparsity, including subword segmentation, back translation, and the copied corpus method. We evaluate the models using the BLEU score and find that transformer outperforms the classic seq2seq model in all settings, in addition to being less sensitive to noise. In general, the best scores are achieved when training the models on word-level-based units. For subword-level models, using back translation proves to be slightly beneficial in low-resource (WO) to high-resource (FR) language translation for the transformer (but not for the seq2seq) models. A slight improvement can also be observed when injecting copied monolingual text in the target language. Moreover, combining the copied method data with back translation leads to a substantial improvement of the translation quality.

Keywords: backtranslation, low-resource language, neural machine translation, sequence-to-sequence, transformer, Wolof

Procedia PDF Downloads 112

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 102

Authors: Soumia Hammadi, Imane Nouacer, Lamine Hamida, Younes A. Hammadi, Rachid Chaibi

Abstract:

Aims and objective: In the present work, we investigate the impact of both acute and chronic diabetes mellitus induced by streptozotocin (STZ) on the hypothalamus of the small gerbil (Gerbillus gerbillus). In this purpose, we aimed to study the histologic structure of the gerbil’s hypothalamic supraoptic (NSO) and paraventricular nucleus (NPV) at two distinct time points: two days and 30 days after diabetes onset. Methods: We conducted our investigation using 19 adult male gerbils weighing 25 to 28 g, divided into three groups as follow: Group I: Control gerbils (n=6) received an intraperitoneal injection of citrate buffer. Group II: STZ-diabetic gerbils (n=8) received a single intraperitoneal injection of STZ at a dose of 165 mg/kg of body weight. Diabetes onset (D0) is considered with the first hyperglycemia level exceeding 2,5 g/L. This group was further divided into two subgroups: Group II-1: Experimental Gerbils, at acute state of diabetes (n=8) sacrificed after 02 days of diabetes onset, Group II-2: Experimental Gerbils at chronic state of diabetes (n=7) sacrificed after 30 days of diabetes onset. Two and 30 days after diabetes onset, gerbils had blood drawn from the retro-orbital sinus into EDTA tubes. After centrifugation at -4°C, plasma was frozen at -80°C for later measurement of Cortisol, ACTH, and insulin. Afterward, animals were decapitated; their brain was removed, weighed, fixed in aqueous bouin, and processed and stained with Toluidine Bleu stain for histo-stereological analysis. A comparison was done with control gerbils treated with citrate buffer. Results: Compared to control gerbils, at 02 Days post diabetes onset, the neuronal somata of the paraventricular (NPV) and supraoptic nuclei (NSO) expressed numerous vacuoles of various sizes, we distinct also a neuronal juxtaposition and several unidentifiable vacuolated profiles were also seen in the neuropile. At the same time, we revealed the presence of à shrunken and condensed nuclei, which seem to touch the parvocellular neurons ( NPV); this leads us to suggest the presence of an apoptotic process in the early stage of diabetes. At 30 days of diabetes mellitus, the NPV manifests a few neurons with a distant appearance, in addition the magnocellular neurons in both NPV and NSO were hypertrophied with a rich euchromatin nucleus, a well-defined nucleolus, and a granular cytoplasm. Despite the neuronal degeneration at this stage, unexpectedly, ACTH registers a continuous significant high level compared to the early stage of diabetes mellitus and to control gerbils. Conclusion: The results suggest that the induction of diabetes mellitus using STZ in the small gerbils lead to alterations in the structure and morphology of the hypothalamus and hyper-secretion of ACTH and cortisol, possibly indicating hyperactivity of the hypothalamo-pituitary adrenal axis (HPA) during both the early and later stages of the disease. The subsequent quantitative evaluation of CRH, immunehistochemical evaluation of apoptosis, and oxidative stress assessment could corroborate our results.

Keywords: diabetes type 1., streptozotocin., small gerbil., hypothalamus., paraventricular nucleus., supraoptic nucleus.

Procedia PDF Downloads 33