Sentence Generation from Conceptual Graph Using Deep Learning

Here the power of a recurrent neural network (RNN) has been exhibited for generating grammatically correct new text from given input text and translation of the new text to the Hindi language with modified bilingual evaluation understudy (BLEU) metric score. Our system aims to generate a grammatically correct new text from given input sentences or paragraphs and translate generated text to Hindi with high translation score. To accomplish a grammatically correct sentence, natural language toolkit (NLTK) is used for grammar correction at the end of text generation. RNN is not very useful for text generation of a gated connection decided to be used for this purpose. The generated text is transferred to machine translation (MT) module. For MT since evaluation is done by humans is a time-consuming task and results differ from evaluator to another evaluator. Hence, the need for assessment of translation system is emerged. The synonym issue is not considered by the BLEU metric. A synonym is treated as a separate word. A modified BLEU (M-BLEU) has been developed as evaluation metrics. It includes several features such as replacing synonym and shallow modules of parsing. The final score of translation is given by BLEU metric scores. Finally, two outputs are there: one is generated text (English) and second is translated text with an improved translation score (Hindi).

Dual coding of knowledge in the human brain.

Conceptual graphs as a universal knowledge representation

Surrounded by structured data, such as medical data, financial data, knowledge bases, etc., data-to-text generation has become an important natural language processing task that can help people better understand the meaning of those data by providing them with user-friendly text. Existing methods for data-to-text generation show promising results in tackling two major challenges: content planning and surface realization, which transform structured data into fluent text. However, they lack an iterative refinement process for generating text, which can enable the model to perfect the text step-by-step while accepting control over the process. In this paper, we explore enhancing data-to-text generation with an iterative refinement process via diffusion. We have four main contributions: (1) we use the diffusion model to improve the prefix tuning for data-to-text generation; (2) we propose a look-ahead guiding loss to supervise the iterative refinement process for better text generation; (3) we extract content plans from reference text and propose a planning-then-writing pipeline to give the model content planning ability; and (4) we conducted experiments on three data-to-text generation datasets and both automatic evaluation criteria (BLEU, NIST, METEOR, ROUGEL, CIDEr, TER, MoverScore, BLEURT, and BERTScore) and human evaluation criteria (Quality and Naturalness) show the effectiveness of our model. Our model can improve the competitive prefix tuning method by 2.19% in terms of a widely-used automatic evaluation criterion BLEU (BiLingual Evaluation Understudy) on WebNLG dataset with GPT-2 Large as the pretrained language model backbone. Human evaluation criteria also show that our model can improve the quality and naturalness of the generated text across all three datasets.

Style is an integral component of a sentence indicated by the choice of words a person makes. Different people have different ways of expressing themselves; however, they adjust their speaking and writing style to a social context, an audience, an interlocutor, or the formality of an occasion. Text style transfer is defined as a task of adapting and/or changing the stylistic manner in which a sentence is written, while preserving the meaning of the original sentence. A systematic review of text style transfer methodologies using deep learning is presented in this article. We point out the technological advances in deep neural networks that have been the driving force behind current successes in the fields of natural language understanding and generation. The review is structured around two key stages in the text style transfer process, namely, representation learning and sentence generation in a new style. The discussion highlights the commonalities and differences between proposed solutions as well as challenges and opportunities that are expected to direct and foster further research in the field.

This study aims to compare the effectiveness of two popular machine translation systems (Google Translate and Babylon machine translation system) used to translate English sentences into Arabic relative to the effectiveness of English to Arabic human translation. There are many automatic methods used to evaluate different machine translators, one of these methods; Bilingual Evaluation Understudy (BLEU) method, which was adopted and implemented to achieve the main goal of this study. BLEU method is based on the assumptions of automated measures that depend on matching machine translators' output to human reference translations; the higher the score, the closer the translation to the human translation will be. Well known English sayings in addition to manually collected sentences from different Internet web sites were used for evaluation purposes. The results of this study have showed that Google machine translation system is better than Babylon machine translation system in terms of precision of translation from English to Arabic.

Recent major developments in semantic communication systems stem from integration of deep learning (DL) techniques. Following the discovery of capacity achieving codes, the primary motivation for adopting the semantic approach, which retrieves meaning without requiring an exact reconstruction, is its potential to further conserve resources such as bandwidth and power. In this paper, we propose a novel semantic communication framework for textual data over additive white Gaussian noise (AWGN) channels via DL. Our framework leverages the information bottleneck (IB) principle to balance minimizing bit transmission under wireless channel rate constraints with maximizing semantic information retention. Unlike previous works, we integrate the bilingual evaluation understudy (BLEU) sentence similarity score into the training objective to enhance model performance. In particular, inspired by knowledge distillation, we utilize large language models (LLMs) during training to transfer their knowledge of text semantics into our model. Using IB principle, we train a neural semantic encoder at the transmitter and a neural semantic decoder at the receiver that incorporates into its objective function the rate constraint together with the BLEU score and the knowledge encoded in the soft probabilities produced by the LLM. Through extensive experiments, our proposed framework demonstrates a notable improvement of up to 45% in text semantic similarity compared to state-of-the-art benchmarks operating at the same channel capacity, significantly outperforming traditional communication systems. Moreover, it exhibits robustness to variations in signal-to-noise ratio (SNR) and achieves significant gains across both low and medium SNR regimes.

The concept of stratified graph introduces some method of knowledge representation (see [T¸ and ˘ areanu, N., ˘ Knowledge representation by labeled stratified graphs, Proc. 8th World Multi-Conference on Systemics, Cybernetics and Informatics, 5 (2004), 345–350; T¸ and ˘ areanu, N., ˘ Proving the Existence of Labelled Stratified Graphs, An. Univ. Craiova Ser. Mat. Inform., 27 (2000), 81–92]) The inference process developed for this method uses the paths of the stratified graphs, an order between the elementary arcs of a path and some results of universal algebras. The order is defined by considering a structured path instead of a regular path. In this paper we define the concept of system of knowledge representation as a tuple of the following components: a stratified graph G, a partial algebra Y of real objects, an embedding mapping (an injective mapping that embeds the nodes of G into objects of Y ) and a set of algorithms such that each of them can combine two objects of Y to get some other object of Y . We define also the concept of inference process performed by a system of knowledge processing in which the interpretation of the symbolic elements is defined by means of natural language constructions. In this manner we obtained a mechanism for texts generation in a natural language (for this approach, Romanian).

Applying convolutional attention mechanisms and Human Memory Search for effective English-Urdu translation

Regional languages play an important role in cultural preservation and strengthening ethnic identity. Bugis, one of the most widely spoken regional languages, can be studied through speech recognition technology. This study analyzes the performance of two Deep Learning models, Long Short-Term Memory (LSTM) and Bidirectional Long Short-Term Memory (BiLSTM), for converting Bugis speech into text. The dataset consists of 2,500 audio recordings from native speakers along with their transcripts. The preprocessing stage includes trimming, denoising, and feature extraction using Mel-Frequency Cepstral Coefficients (MFCC). Evaluation using Word Error Rate (WER) and Bilingual Evaluation Understudy (BLEU) shows that BiLSTM outperforms LSTM, achieving a WER of 4.88% and a BLEU of 86.82%, compared to LSTM’s WER of 90.33% and BLEU of 3.86%. These findings indicate that BiLSTM is more effective in capturing contextual information, resulting in more accurate transcriptions. This research contributes to the preservation of the Bugis language and provides insights into the application of Deep Learning in low-resource speech recognition. Future work will explore attention mechanisms and larger datasets to further improve model performance

Different word orders will be resulted in during the generation processes of negative sentences in different natural languages. This brings us difficulties when we deal with the negative sentence generation problem. This paper studies natural language generation of negative sentences and the problem of how to handle the different word orders that occur in the generation processes in different natural languages through feature-checking theory of the minimalist paradigm. Without losing generality, we use examples in the languages of modern standard English (MSE), French, and early modern English (EME) to illustrate the issue. The main technical point is that weak tense- and agreement-features can be separated from the word that carries them and percolate up the structural tree on their own to get themselves checked; strong tense- and agreement-features cannot be separated from the word that carries them and can only move together with the phonetic form of the word when they need to get themselves checked. And this leads to different kinds of movement (i.e. attraction and movement). For the parameter of strength, we may set its value as (+) for EME and French, and (-) for MSE. Different parameter values lead to different kinds of movement that can be involved in the negative sentence generation processes. Therefore, the difference in word orders of negative sentences across languages can be handled uniformly through the approach of parameter-setting. By the study carried out in this paper, we might also justify the use of the minimalist program as the grammar that serves in natural language generation from a certain perspective.

Image description creation is utilised in a variety of applications, such as automatic script development, medical image annotation, and news reporting. However, the detailed description of the image with the relevant region is complicated using deep learning approaches. In this paper, an intelligent Red-tailed Hawk Generative Deep Learning Model is presented for generating image descriptions. The suggested technique reduces the loss function of the network model while also updating the weight parameter using the meta-heuristic Red-tailed Hawk Optimizer (RTHO). Extensive simulation experiments are carried out to assess the suggested model. The obtained results demonstrate that utilising the proposed approach improves the BiLingual Evaluation Understudy (BLEU)-1, BLEU-2, and METEOR scores to 0.74, 0.44, and 0.53, respectively.

Conceptual graphs and fuzzy logic are two logical formalisms that emphasize the target of natural language, where conceptual graphs provide a structure of formulas close to that of natural language sentences while fuzzy logic provides a methodology for computing with words. This paper proposes fuzzy conceptual graphs as a knowledge representation language that combines the advantages of both the two formalisms for artificial intelligence approaching human expression and reasoning. Firstly, the conceptual graph language is extended with functional relation types for representing functional dependency, and conjunctive types for joining concepts and relations. Then fuzzy conceptual graphs are formulated as a generalization of conceptual graphs where fuzzy types and fuzzy attribute-values are used in place of crisp types and crisp attribute-values. Projection and join as basic operations for reasoning on fuzzy conceptual graphs are defined, taking into account the semantics of fuzzy set-based values.

In this paper, the generation component of the SUSY machine translation system is described. In doing this, the author attempts to explain some basic differences in natural language generation in MT and AI. This more general discussion concentrates on the aspect of syntactic and/or semantic representation, especially on the status of deep semantic cases in natural language processing. The presentation of the SUSY generation component includes detailed descriptions of the data base for generation and its manipulation as well as the generation process and sane linguistic aspects of generation.

This paper describes two approaches to the automatic generation of behavioral VHDL models from descriptions written in natural language. Both approaches are based on a modeling style in which behavior is represented by a system of interconnected processes. The first approach employs a semantic grammar to directly generate a single VHDL process from a paragraph written in a restricted English called ModelSpeak. The second approach accepts more general English and generates models consisting of multiple processes. series of sentences describing the behavior of a single process and generated correct VHDL code for it. This interface requires the behavior be expressed in a highly restricted English called ModelSpeak (10). Conditional expressions (if statements) may be nested to two levels, and interaction is invoked by the interface to resolve scope ambiguity of loops. This interface is implemented with a semantic grammar (3) in Prolog. The second interface (Approach II) to the Modeler's Assistant for natural language is substantially more complex, but accepts more general English expressions and generates multiple-process models. In this approach, each sentence is parsed by a bottom-up, parallel chart parser to generate a set of valid parse trees. The trees are examined by a Semantic Analyzer (7) to determine one which is semantically correct, and a conceptual graph (14) (semantic network) of its meaning is generated. Once the conceptual graph of individual sentences are joined together (12), a program called the Linker (8) analyzes the graph to produce a Process Model Graph and VHDL process code for the Modeler's Assistant. Nodes in the conceptual graph representing verbs typically generate processes or the conditions which govern them. Nodes representing objects generate signals. The modeling style underlying the Linker produces a sizable collection of processes, each consisting of a single if_then statement. To avoid unnecessary bus resolutions, multiple processes assigning to the same signal are merged. Since signal types are generally not expressed in the English description, they may be added interactively later. Before describing the approaches in more detail, related research on natural language interfaces is reviewed briefly.