SearcharxivSearch

arXiv subjects

Chiranjib Sur

Publications and source records attributed to Chiranjib Sur.

At least 19 recordsLinked to original sources

Learning to Extract Cross-Domain Aspects and Understanding Sentiments Using Large Language Models

Aspect-based sentiment analysis (ASBA) is a refined approach to sentiment analysis that aims to extract and classify sentiments based on specific aspects or features of a product, service, or entity. Unlike traditional sentiment analysis, which assigns a general sentiment score to entire reviews or texts, ABSA focuses on breaking down the text into individual components or aspects (e.g., quality, price, service) and evaluating the sentiment towards each. This allows for a more granular level of understanding of customer opinions, enabling businesses to pinpoint specific areas of strength and improvement. The process involves several key steps, including aspect extraction, sentiment classification, and aspect-level sentiment aggregation for a review paragraph or any other form that the users have provided. ABSA has significant applications in areas such as product reviews, social media monitoring, customer feedback analysis, and market research. By leveraging techniques from natural language processing (NLP) and machine learning, ABSA facilitates the extraction of valuable insights, enabling companies to make data-driven decisions that enhance customer satisfaction and optimize offerings. As ABSA evolves, it holds the potential to greatly improve personalized customer experiences by providing a deeper understanding of sentiment across various product aspects. In this work, we have analyzed the strength of LLMs for a complete cross-domain aspect-based sentiment analysis with the aim of defining the framework for certain products and using it for other similar situations. We argue that it is possible to that at an effectiveness of 92\% accuracy for the Aspect Based Sentiment Analysis dataset of SemEval-2015 Task 12.

cs.CL

SACT: Self-Aware Multi-Space Feature Composition Transformer for Multinomial Attention for Video Captioning

Video captioning works on the two fundamental concepts, feature detection and feature composition. While modern day transformers are beneficial in composing features, they lack the fundamental problems of selecting and understanding of the contents. As the feature length increases, it becomes increasingly important to include provisions for improved capturing of the pertinent contents. In this work, we have introduced a new concept of Self-Aware Composition Transformer (SACT) that is capable of generating Multinomial Attention (MultAtt) which is a way of generating distributions of various combinations of frames. Also, multi-head attention transformer works on the principle of combining all possible contents for attention, which is good for natural language classification, but has limitations for video captioning. Video contents have repetitions and require parsing of important contents for better content composition. In this work, we have introduced SACT for more selective attention and combined them for different attention heads for better capturing of the usable contents for any applications. To address the problem of diversification and encourage selective utilization, we propose the Self-Aware Composition Transformer model for dense video captioning and apply the technique on two benchmark datasets like ActivityNet and YouCookII.

cs.CV

Self-Segregating and Coordinated-Segregating Transformer for Focused Deep Multi-Modular Network for Visual Question Answering

Attention mechanism has gained huge popularity due to its effectiveness in achieving high accuracy in different domains. But attention is opportunistic and is not justified by the content or usability of the content. Transformer like structure creates all/any possible attention(s). We define segregating strategies that can prioritize the contents for the applications for enhancement of performance. We defined two strategies: Self-Segregating Transformer (SST) and Coordinated-Segregating Transformer (CST) and used it to solve visual question answering application. Self-segregation strategy for attention contributes in better understanding and filtering the information that can be most helpful for answering the question and create diversity of visual-reasoning for attention. This work can easily be used in many other applications that involve repetition and multiple frames of features and would reduce the commonality of the attentions to a great extent. Visual Question Answering (VQA) requires understanding and coordination of both images and textual interpretations. Experiments demonstrate that segregation strategies for cascaded multi-head transformer attention outperforms many previous works and achieved considerable improvement for VQA-v2 dataset benchmark.

cs.CV

ReLGAN: Generalization of Consistency for GAN with Disjoint Constraints and Relative Learning of Generative Processes for Multiple Transformation Learning

Image to image transformation has gained popularity from different research communities due to its enormous impact on different applications, including medical. In this work, we have introduced a generalized scheme for consistency for GAN architectures with two new concepts of Transformation Learning (TL) and Relative Learning (ReL) for enhanced learning image transformations. Consistency for GAN architectures suffered from inadequate constraints and failed to learn multiple and multi-modal transformations, which is inevitable for many medical applications. The main drawback is that it focused on creating an intermediate and workable hybrid, which is not permissible for the medical applications which focus on minute details. Another drawback is the weak interrelation between the two learning phases and TL and ReL have introduced improved coordination among them. We have demonstrated the capability of the novel network framework on public datasets. We emphasized that our novel architecture produced an improved neural image transformation version for the image, which is more acceptable to the medical community. Experiments and results demonstrated the effectiveness of our framework with enhancement compared to the previous works.

cs.CV

Gaussian Smoothen Semantic Features (GSSF) -- Exploring the Linguistic Aspects of Visual Captioning in Indian Languages (Bengali) Using MSCOCO Framework

In this work, we have introduced Gaussian Smoothen Semantic Features (GSSF) for Better Semantic Selection for Indian regional language-based image captioning and introduced a procedure where we used the existing translation and English crowd-sourced sentences for training. We have shown that this architecture is a promising alternative source, where there is a crunch in resources. Our main contribution of this work is the development of deep learning architectures for the Bengali language (is the fifth widely spoken language in the world) with a completely different grammar and language attributes. We have shown that these are working well for complex applications like language generation from image contexts and can diversify the representation through introducing constraints, more extensive features, and unique feature spaces. We also established that we could achieve absolute precision and diversity when we use smoothened semantic tensor with the traditional LSTM and feature decomposition networks. With better learning architecture, we succeeded in establishing an automated algorithm and assessment procedure that can help in the evaluation of competent applications without the requirement for expertise and human intervention.

cs.CL

MRRC: Multiple Role Representation Crossover Interpretation for Image Captioning With R-CNN Feature Distribution Composition (FDC)

While image captioning through machines requires structured learning and basis for interpretation, improvement requires multiple context understanding and processing in a meaningful way. This research will provide a novel concept for context combination and will impact many applications to deal visual features as an equivalence of descriptions of objects, activities and events. There are three components of our architecture: Feature Distribution Composition (FDC) Layer Attention, Multiple Role Representation Crossover (MRRC) Attention Layer and the Language Decoder. FDC Layer Attention helps in generating the weighted attention from RCNN features, MRRC Attention Layer acts as intermediate representation processing and helps in generating the next word attention, while Language Decoder helps in estimation of the likelihood for the next probable word in the sentence. We demonstrated effectiveness of FDC, MRRC, regional object feature attention and reinforcement learning for effective learning to generate better captions from images. The performance of our model enhanced previous performances by 35.3\% and created a new standard and theory for representation generation based on logic, better interpretability and contexts.

cs.LG

aiTPR: Attribute Interaction-Tensor Product Representation for Image Caption

Region visual features enhance the generative capability of the machines based on features, however they lack proper interaction attentional perceptions and thus ends up with biased or uncorrelated sentences or pieces of misinformation. In this work, we propose Attribute Interaction-Tensor Product Representation (aiTPR) which is a convenient way of gathering more information through orthogonal combination and learning the interactions as physical entities (tensors) and improving the captions. Compared to previous works, where features are added up to undefined feature spaces, TPR helps in maintaining sanity in combinations and orthogonality helps in defining familiar spaces. We have introduced a new concept layer that defines the objects and also their interactions that can play a crucial role in determination of different descriptions. The interaction portions have contributed heavily for better caption quality and has out-performed different previous works on this domain and MSCOCO dataset. We introduced, for the first time, the notion of combining regional image features and abstracted interaction likelihood embedding for image captioning.

cs.NE

TPsgtR: Neural-Symbolic Tensor Product Scene-Graph-Triplet Representation for Image Captioning

Image captioning can be improved if the structure of the graphical representations can be formulated with conceptual positional binding. In this work, we have introduced a novel technique for caption generation using the neural-symbolic encoding of the scene-graphs, derived from regional visual information of the images and we call it Tensor Product Scene-Graph-Triplet Representation (TP$_{sgt}$R). While, most of the previous works concentrated on identification of the object features in images, we introduce a neuro-symbolic embedding that can embed identified relationships among different regions of the image into concrete forms, instead of relying on the model to compose for any/all combinations. These neural symbolic representation helps in better definition of the neural symbolic space for neuro-symbolic attention and can be transformed to better captions. With this approach, we introduced two novel architectures (TP$_{sgt}$R-TDBU and TP$_{sgt}$R-sTDBU) for comparison and experiment result demonstrates that our approaches outperformed the other models, and generated captions are more comprehensive and natural.

cs.CL

CRUR: Coupled-Recurrent Unit for Unification, Conceptualization and Context Capture for Language Representation -- A Generalization of Bi Directional LSTM

In this work we have analyzed a novel concept of sequential binding based learning capable network based on the coupling of recurrent units with Bayesian prior definition. The coupling structure encodes to generate efficient tensor representations that can be decoded to generate efficient sentences and can describe certain events. These descriptions are derived from structural representations of visual features of images and media. An elaborated study of the different types of coupling recurrent structures are studied and some insights of their performance are provided. Supervised learning performance for natural language processing is judged based on statistical evaluations, however, the truth is perspective, and in this case the qualitative evaluations reveal the real capability of the different architectural strengths and variations. Bayesian prior definition of different embedding helps in better characterization of the sentences based on the natural language structure related to parts of speech and other semantic level categorization in a form which is machine interpret-able and inherits the characteristics of the Tensor Representation binding and unbinding based on the mutually orthogonality. Our approach has surpassed some of the existing basic works related to image captioning.

cs.CL

Feature Fusion Effects of Tensor Product Representation on (De)Compositional Network for Caption Generation for Images

Progress in image captioning is gradually getting complex as researchers try to generalized the model and define the representation between visual features and natural language processing. This work tried to define such kind of relationship in the form of representation called Tensor Product Representation (TPR) which generalized the scheme of language modeling and structuring the linguistic attributes (related to grammar and parts of speech of language) which will provide a much better structure and grammatically correct sentence. TPR enables better and unique representation and structuring of the feature space and will enable better sentence composition from these representations. A large part of the different ways of defining and improving these TPR are discussed and their performance with respect to the traditional procedures and feature representations are evaluated for image captioning application. The new models achieved considerable improvement than the corresponding previous architectures.

cs.CV

Green Heron Swarm Optimization Algorithm - State-of-the-Art of a New Nature Inspired Discrete Meta-Heuristics

Many real world problems are NP-Hard problems are a very large part of them can be represented as graph based problems. This makes graph theory a very important and prevalent field of study. In this work a new bio-inspired meta-heuristics called Green Heron Swarm Optimization (GHOSA) Algorithm is being introduced which is inspired by the fishing skills of the bird. The algorithm basically suited for graph based problems like combinatorial optimization etc. However introduction of an adaptive mathematical variation operator called Location Based Neighbour Influenced Variation (LBNIV) makes it suitable for high dimensional continuous domain problems. The new algorithm is being operated on the traditional benchmark equations and the results are compared with Genetic Algorithm and Particle Swarm Optimization. The algorithm is also operated on Travelling Salesman Problem, Quadratic Assignment Problem, Knapsack Problem dataset. The procedure to operate the algorithm on the Resource Constraint Shortest Path and road network optimization is also discussed. The results clearly demarcates the GHOSA algorithm as an efficient algorithm specially considering that the number of algorithms for the discrete optimization is very low and robust and more explorative algorithm is required in this age of social networking and mostly graph based problem scenarios.

cs.NE

A relativistic unitary coupled-cluster study of electric quadrupole moment and magnetic dipole hyperfine constants of ^{199}Hg^{+}

Searching for an accurate optical clock which can serve as a better time standard than the present day atomic clock is highly demanding from several areas of science and technology. Several attempts have been made to built more accurate clocks with different ion species. In this article we discuss the electric quadrupole and hyperfine shifts in the $5d^{9}6s^{2} ^{2}D_{5/2}(F=0,m_{F}=0)\leftrightarrow5d^{10}6s ^{2}S_{1/2}(F=2,m_{F}=0)$ clock transition in $ ^{199}Hg^{+}$, one of the most promising candidates for next generation optical clocks. We have applied Fock-space unitary coupled-cluster (FSUCC) theory to study the electric quadrupole moment of the $5d^{9}6s^{2} ^{2}D_{5/2}$ state and magnetic dipole hyperfine constants of $5d^{9}6s^{2} ^{2}D_{3/2,5/2}$ and $5d^{10}6s^{1} ^{2}S_{1/2}$ states respectively of $ ^{199}Hg^{+}$. We have also compared our results with available data. To the best of our knowledge, this is the first time a variant of coupled-cluster (CC) theories has been applied to study these kinds of properties of $Hg^{+}$and is the most accurate estimate of these quantities to date.

physics.atom-ph

Effects of partial triple excitations in atomic coupled cluster calculations

In this article we study the effects of higher body excitations in the relativistic CC calculations for atoms and ions with one valence electron using Fock-space CCSD, CCSD(T) and its unitary variants. The present study demonstrates that CCSD(T) estimates the ionization potentials (IPs) and the valence electron removal energies quite accurately for alkali atoms and singly ionized alkaline earth ions, but yields unphysical energy levels for atoms and/or ions with partially filled sub-shell like C II. We further demonstrate that the higher body excitation effects can be incorporated more effectively through the unitary coupled cluster theory (UCC) compared to the CCSD(T) method.

physics.atom-ph

Branching ratios of radiative transitions in O VI

We study the branching ratios of the allowed and forbidden radiative transitions among the first few (9) fine structure levels of O VI using relativistic coupled cluster theory. We find irregular patterns for a number of transitions with in $n$-complexes with $n\le4$. We have used the exisiting values of the allowed electric dipole ($E1$) transition as a benchmark of our theory. Good agreement with the existing values establish accuracies of not only the theoretical method but the basis function as well. In general the electric quadrupole ($E2$) transition probabilities are greater in magnitude than magnetic dipole ($M1$) transition probabilities, whereas for medium atomic transition frequencies they are of the same order of magnitude. On the other hand if the transitions involved are between two fine structure components of the same term, then the $M1$ transition probability is more probable than that of $E2$. We have analyzed these trends with physical arguments and order of magnitude estimations. The results presented here in tabular and graphical forms are compared with the available theoretical and observed data. Graphical analysis helps to understand the trends of electric and magnetic transitions for the decay channels presented here. Our calculated values of the lifetimes of the excited states are in very good agreement with the available results.

physics.atom-ph

Relativistic multi-reference Fock-space coupled-cluster calculation of the forbidden $6s^2^1 S_0 \longrightarrow 6s5d^3 D_1$ magnetic-dipole transition in ytterbium

We report the forbidden $6s^{2} ^{1}S_{0}\longrightarrow6s5d ^{3}D_{1}$ magnetic-dipole transition amplitude computed using multi-reference Fock-space coupled-cluster theory. Our computed transition matrix element ($1.34\times10^{-4}μ_{B}$) is in excellent agreement with the experimental value ($1.33\times10^{-4}$ $μ_{B}$). This value in combination with other known quantities will be helpful to determine the parity non-conserving amplitude for the $6s^{2} ^{1}S_{0}\longrightarrow6s5d ^{3}D_{1}$ transition in atomic Yb. To our knowledge our calculation is the most accurate to date and can be very important in the search of physics beyond the standard model. We further report the $6s6p ^{3}P_{0}\longrightarrow6s6p ^{1}P_{1}$ and $6s5d ^{3}D_{1}\longrightarrow6s6p ^{3}P_{0}$ transition matrix elements which are also in good agreement with the earlier theoretical estimates.

physics.atom-ph

Optical frequency standard with $Sr^+$: A theoretical many-body approach

Demands from several areas of science and technology have lead to a worldwide search for accurate optical clocks with an uncertainty of 1 part in $10^{18}$, which is $10^{3}$ times more accurate than the present day cesium atomic clocks based on microwave frequency regime. In this article we discuss the electric quadrupole and the hyperfine shifts in the $5s ^{2}S_{1/2}\longrightarrow4d ^{2}D_{5/2}$ clock transition in $\mathrm{Sr^{+}}$, one of the most promising candidates for next generation optical clocks. We have applied relativistic coupled cluster theory for determining the electric quadrupole moment of the $4d ^{2}D_{5/2}$ state of $\mathrm{^{88}Sr^{+}}$ and the magnetic dipole ($A$) and electric quadrupole ($B$) hyperfine constants for the $5s ^{2}S_{1/2}$ and $4d ^{2}D_{5/2}$ states which are important in the study of frequency standards with $\mathrm{Sr^{+}}$. The effects of electron correlation which are very crucial for the accurate determination of these quantities have been discussed.

physics.atom-ph

Random Phase Approximation For allowed and Parity Non-conserving Electric Dipole Transition Amplitudes and its connection with Many-Body Perturbation Theory and Coupled Cluster Theory

The connections between the Random Phase Approximation (RPA) and Many-Body Perturbation Theory (MBPT) and its all order generalisation, the Coupled- Cluster Theory (CCT) have been explored. Explicit expressions have been derived for the electric dipole amplitudes for allowed and forbidden transitions induced by the parity non-conserving neutral weak interaction. The Goldstone diagrams associated with the RPA terms in both cases are shown to arise in MBPT and CCT and the numerical verification of this relationship is made for the allowed electric dipole transitions.

physics.atom-ph

Electric quadrupole moment of the $4d ^{2}D_{5/2}$ state in $^{88}Sr^{+}$ and its role in an optical frequency standard

The electric quadrupole moment for the $4d ^2D_{5/2}$ state of $\mathrm{^{88}Sr^+}$, one of the most important candidates for an optical clock, has been calculated using the relativistic coupled-cluster theory. The result of the calculation is presented and the important many-body contributions are highlighted. The calculated electric quadrupole moment is $(2.94\pm0.07)ea_{0}^{2}$, where $a_{0}$ is the Bohr radius and $e$ the electronic charge while the measured value is $(2.6\pm0.3)ea_{0}^{2}$. This is so far the most accurate determination of the electric quadrupole moment for the above mentioned state.

physics.atom-ph