arXiv subjects

Saptarshi Ghosh

Publications and source records attributed to Saptarshi Ghosh.

At least 91 records · Page 5Linked to original sources

arXiv2021

An Unsupervised Normalization Algorithm for Noisy Text: A Case Study for Information Retrieval and Stance Detection

A large fraction of textual data available today contains various types of 'noise', such as OCR noise in digitized documents, noise due to informal writing style of users on microblogging sites, and so on. To enable tasks such as search/retrieval and classification over all the available data, we need robust algorithms for text normalization, i.e., for cleaning different kinds of noise in the text. There have been several efforts towards cleaning or normalizing noisy text; however, many of the existing text normalization methods are supervised and require language-dependent resources or large amounts of training data that is difficult to obtain. We propose an unsupervised algorithm for text normalization that does not need any training data / human intervention. The proposed algorithm is applicable to text over different languages, and can handle both machine-generated and human-generated noise. Experiments over several standard datasets show that text normalization through the proposed algorithm enables better retrieval and stance detection, as compared to that using several baseline text normalization methods. Implementation of our algorithm can be found at https://github.com/ranarag/UnsupClean.

cs.IR↗

arXiv2020

ZSCRGAN: A GAN-based Expectation Maximization Model for Zero-Shot Retrieval of Images from Textual Descriptions

Most existing algorithms for cross-modal Information Retrieval are based on a supervised train-test setup, where a model learns to align the mode of the query (e.g., text) to the mode of the documents (e.g., images) from a given training set. Such a setup assumes that the training set contains an exhaustive representation of all possible classes of queries. In reality, a retrieval model may need to be deployed on previously unseen classes, which implies a zero-shot IR setup. In this paper, we propose a novel GAN-based model for zero-shot text to image retrieval. When given a textual description as the query, our model can retrieve relevant images in a zero-shot setup. The proposed model is trained using an Expectation-Maximization framework. Experiments on multiple benchmark datasets show that our proposed model comfortably outperforms several state-of-the-art zero-shot text to image retrieval models, as well as zero-shot classification and hashing models suitably used for retrieval.

cs.CV↗

arXiv2020

A Survey on Disaster: Understanding the After-effects of Super-cyclone Amphan and Helping Hand of Social Media

The super-cyclonic storm "Amphan" hit Eastern India, specifically the state of West Bengal, Odisha and parts of Bangladesh in May 2020, and caused severe damage to the regions. In this study, we aim to understand the self-reported effects of this natural disaster on residents of the state of West Bengal. To that end, we conducted an online survey to understand the effects of the cyclone. In total, 201 participants (spanning five districts) from the worst-affected state of West Bengal participated in the survey. This report describes our findings from the survey, with respect to the damages caused by the cyclone, how it affected the population in various districts of West Bengal, and how prepared the authorities were in responding to the disaster. We found that the participants were most adversely affected in this disaster due to disruption of services like electricity, phone and internet (as opposed to uprooting of trees and water-logging). Furthermore, we found that receiving responses to Amphan-related queries is highly positively correlated with the favorable perception of people about preparedness of authorities. Additionally, we study the usage of online social media by the affected population in the days immediately after the disaster. Our results strongly suggest how social media platforms can help authorities to better prepare for future disasters. In summary, our study analyzes self-reported data collected from grassroots, and brings out several key insights that can help authorities deal better with disaster events in future.

cs.CY↗

arXiv2020

Utilizing Microblogs for Assisting Post-Disaster Relief Operations via Matching Resource Needs and Availabilities

During a disaster event, two types of information that are especially useful for coordinating relief operations are needs and availabilities of resources (e.g., food, water, medicines) in the affected region. Information posted on microblogging sites is increasingly being used for assisting post-disaster relief operations. In this context, two practical challenges are (i)~to identify tweets that inform about resource needs and availabilities (termed as need-tweets and availability-tweets respectively), and (ii)~to automatically match needs with appropriate availabilities. While several works have addressed the first problem, there has been little work on automatically matching needs with availabilities. The few prior works that attempted matching only considered the resources, and no attempt has been made to understand other aspects of needs/availabilities that are essential for matching in practice. In this work, we develop a methodology for understanding five important aspects of need-tweets and availability-tweets, including what resource and what quantity is needed/available, the geographical location of the need/availability, and who needs / is providing the resource. Understanding these aspects helps us to address the need-availability matching problem considering not only the resources, but also other factors such as the geographical proximity between the need and the availability. To our knowledge, this study is the first attempt to develop methods for understanding the semantics of need-tweets and availability-tweets. We also develop a novel methodology for matching need-tweets with availability-tweets, considering both resource similarity and geographical proximity. Experiments on two datasets corresponding to two disaster events, demonstrate that our proposed methods perform substantially better matching than those in prior works.

cs.SI↗

arXiv2020

Stance Detection in Web and Social Media: A Comparative Study

Online forums and social media platforms are increasingly being used to discuss topics of varying polarities where different people take different stances. Several methodologies for automatic stance detection from text have been proposed in literature. To our knowledge, there has not been any systematic investigation towards their reproducibility, and their comparative performances. In this work, we explore the reproducibility of several existing stance detection models, including both neural models and classical classifier-based models. Through experiments on two datasets -- (i)~the popular SemEval microblog dataset, and (ii)~a set of health-related online news articles -- we also perform a detailed comparative analysis of various methods and explore their shortcomings. Implementations of all algorithms discussed in this paper are available at https://github.com/prajwal1210/Stance-Detection-in-Web-and-Social-Media.

cs.CL↗

arXiv2020

Hier-SPCNet: A Legal Statute Hierarchy-based Heterogeneous Network for Computing Legal Case Document Similarity

Computing similarity between two legal case documents is an important and challenging task in Legal IR, for which text-based and network-based measures have been proposed in literature. All prior network-based similarity methods considered a precedent citation network among case documents only (PCNet). However, this approach misses an important source of legal knowledge -- the hierarchy of legal statutes that are applicable in a given legal jurisdiction (e.g., country). We propose to augment the PCNet with the hierarchy of legal statutes, to form a heterogeneous network Hier-SPCNet, having citation links between case documents and statutes, as well as citation and hierarchy links among the statutes. Experiments over a set of Indian Supreme Court case documents show that our proposed heterogeneous network enables significantly better document similarity estimation, as compared to existing approaches using PCNet. We also show that the proposed network-based method can complement text-based measures for better estimation of legal document similarity.

cs.IR↗

arXiv2020

NARMADA: Need and Available Resource Managing Assistant for Disasters and Adversities

Although a lot of research has been done on utilising Online Social Media during disasters, there exists no system for a specific task that is critical in a post-disaster scenario -- identifying resource-needs and resource-availabilities in the disaster-affected region, coupled with their subsequent matching. To this end, we present NARMADA, a semi-automated platform which leverages the crowd-sourced information from social media posts for assisting post-disaster relief coordination efforts. The system employs Natural Language Processing and Information Retrieval techniques for identifying resource-needs and resource-availabilities from microblogs, extracting resources from the posts, and also matching the needs to suitable availabilities. The system is thus capable of facilitating the judicious management of resources during post-disaster relief operations.

cs.SI↗

arXiv2020

Identification of Chimera using Machine Learning

Chimera state refers to coexistence of coherent and non-coherent phases in identically coupled dynamical units found in various complex dynamical systems. Identification of Chimera, on one hand is essential due to its applicability in various areas including neuroscience, and on other hand is challenging due to its widely varied appearance in different systems and the peculiar nature of its profile. Therefore, a simple yet universal method for its identification remains an open problem. Here, we present a very distinctive approach using machine learning techniques to characterize different dynamical phases and identify the chimera state from given spatial profiles generated using various different models. The experimental results show that the performance of the classification algorithms varies for different dynamical models. The machine learning algorithms, namely random forest, oblique random forest based on tikhonov, parallel-axis split and null space regularization achieved more than $96\% $ accuracy for the Kuramoto model. For the logistic-maps, random forest and tikhonov regularization based oblique random forest showed more than $90\%$ accuracy, and for the Hénon-Map model, random forest, null-space and axis-parallel split regularization based oblique random forest achieved more than $80\%$ accuracy. The oblique random forest with null space regularization achieved consistent performance (more than $83\%$ accuracy) across different dynamical models while the auto-encoder based random vector functional link neural network showed relatively lower performance. This work provides a direction for employing machine learning techniques to identify dynamical patterns arising in coupled non-linear units on large-scale, and for characterizing complex spatio-temporal patterns in real-world systems for various applications.

nlin.AO↗

arXiv2020

Methods for Computing Legal Document Similarity: A Comparative Study

Computing similarity between two legal documents is an important and challenging task in the domain of Legal Information Retrieval. Finding similar legal documents has many applications in downstream tasks, including prior-case retrieval, recommendation of legal articles, and so on. Prior works have proposed two broad ways of measuring similarity between legal documents - analyzing the precedent citation network, and measuring similarity based on textual content similarity measures. But there has not been a comprehensive comparison of these existing methods on a common platform. In this paper, we perform the first systematic analysis of the existing methods. In addition, we explore two promising new similarity computation methods - one text-based and the other based on network embeddings, which have not been considered till now.

cs.SI↗

arXiv2019

Identification of Rhetorical Roles of Sentences in Indian Legal Judgments

Automatically understanding the rhetorical roles of sentences in a legal case judgement is an important problem to solve, since it can help in several downstream tasks like summarization of legal judgments, legal search, and so on. The task is challenging since legal case documents are usually not well-structured, and these rhetorical roles may be subjective (as evident from variation of opinions between legal experts). In this paper, we address this task for judgments from the Supreme Court of India. We label sentences in 50 documents using multiple human annotators, and perform an extensive analysis of the human-assigned labels. We also attempt automatic identification of the rhetorical roles of sentences. While prior approaches towards this task used Conditional Random Fields over manually handcrafted features, we explore the use of deep neural models which do not require hand-crafting of features. Experiments show that neural models perform much better in this task than baseline methods which use handcrafted features.

cs.IR↗

arXiv2019

Summarizing User-generated Textual Content: Motivation and Methods for Fairness in Algorithmic Summaries

As the amount of user-generated textual content grows rapidly, text summarization algorithms are increasingly being used to provide users a quick overview of the information content. Traditionally, summarization algorithms have been evaluated only based on how well they match human-written summaries (e.g. as measured by ROUGE scores). In this work, we propose to evaluate summarization algorithms from a completely new perspective that is important when the user-generated data to be summarized comes from different socially salient user groups, e.g. men or women, Caucasians or African-Americans, or different political groups (Republicans or Democrats). In such cases, we check whether the generated summaries fairly represent these different social groups. Specifically, considering that an extractive summarization algorithm selects a subset of the textual units (e.g. microblogs) in the original data for inclusion in the summary, we investigate whether this selection is fair or not. Our experiments over real-world microblog datasets show that existing summarization algorithms often represent the socially salient user-groups very differently compared to their distributions in the original data. More importantly, some groups are frequently under-represented in the generated summaries, and hence get far less exposure than what they would have obtained in the original data. To reduce such adverse impacts, we propose novel fairness-preserving summarization algorithms which produce high-quality summaries while ensuring fairness among various groups. To our knowledge, this is the first attempt to produce fair text summarization, and is likely to open up an interesting research direction.

cs.IR↗

arXiv2019

Delay engineered solitary states in complex networks

We present a technique to engineer solitary states by means of delayed links in a network of neural oscillators and in coupled chaotic maps. Solitary states are intriguing partial synchronization patterns, where a synchronized cluster coexists with solitary nodes displaced from this cluster and distributed randomly over the network. We induce solitary states in the originally synchronized network of identical nodes by introducing delays in the links for a certain number of selected network elements. It is shown that the extent of displacement and the position of solitary elements can be completely controlled by the choice (values) and positions (locations) of the incorporated delays, reshaping the delay engineered solitary states in the network.

nlin.AO↗

arXiv2019

Taming Chimeras in Networks through Multiplexing Delays

Chimera referring to a coexistence of coherent and incoherent states, is traditionally very difficult to control due to its peculiar nature. Here, we provide a recipe to construct chimera states in the multiplex networks with the aid of multiplexing-delays. The chimera state in multiplex networks is produced by introducing heterogeneous delays in a fraction of inter-layer links, referred as multiplexing-delay, in a sequence. Additionally, the emergence of the incoherence in the chimera state can be regulated by making appropriate choice of both inter- and intra-layer coupling strengths, whereas the extent and the position of the incoherence regime can be regulated by appropriate placing and {strength} of the multiplexing delays. The proposed technique to construct such {engineered} chimera equips us with multiplex network's structural parameters as tools in gaining both qualitative- and quantitative-control over the incoherent section of the chimera states and, in turn, the chimera. Our investigation can be of worth in controlling dynamics of multi-level delayed systems and attain desired chimeric patterns.

nlin.AO↗

arXiv2019

Nano radiation detector: A possibility Making Physically Useful Nanostructures With Focused Ion Beam: A study of Nano Gaps

Child-Langmuir law expresses the IV relation between two parallel plates. It has been found that in reality the ordinary Child Langmuir Law does not hold true. It shows some geometric dependence on the system.In this report capacitors of various thickness and septation has been studied both in macro and nano scale to propose a geometrical correction to the law. More over from the studies, a new type of radiation detector was prepared which can detect radiation in small volumes as compared with normal radiation detectors.

physics.ins-det↗

arXiv2019

A Network-centric Framework for Auditing Recommendation Systems

To improve the experience of consumers, all social media, commerce and entertainment sites deploy Recommendation Systems (RSs) that aim to help users locate interesting content. These RSs are black-boxes - the way a chunk of information is filtered out and served to a user from a large information base is mostly opaque. No one except the parent company generally has access to the entire information required for auditing these systems - neither the details of the algorithm nor the user-item interactions are ever made publicly available for third-party auditors. Hence auditing RSs remains an important challenge, especially with the recent concerns about how RSs are affecting the views of the society at large with new technical jargons like "echo chambers", "confirmation biases", "filter bubbles" etc. in place. Many prior works have evaluated different properties of RSs such as diversity, novelty, etc. However, most of these have focused on evaluating static snapshots of RSs. Today, auditors are not only interested in these static evaluations on a snapshot of the system, but also interested in how these systems are affecting the society in course of time. In this work, we propose a novel network-centric framework which is not only able to quantify various static properties of RSs, but also is able to quantify dynamic properties such as how likely RSs are to lead to polarization or segregation of information among their users. We apply the framework to several popular movie RSs to demonstrate its utility.

cs.IR↗

arXiv2018

Ensemble-based Overlapping Community Detection using Disjoint Community Structures

While there has been a plethora of approaches for detecting disjoint communities from real-world complex networks, some methods for detecting overlapping community structures have also been recently proposed. In this work, we argue that, instead of developing separate approaches for detecting overlapping communities, a promising alternative is to infer the overlapping communities from multiple disjoint community structures. We propose an ensemble-based approach, called EnCoD, that leverages the solutions produced by various disjoint community detection algorithms to discover the overlapping community structure. Specifically, EnCoD generates a feature vector for each vertex from the results of the base algorithms and learns which features lead to detect densely connected overlapping regions in an unsupervised way. It keeps on iterating until the likelihood of each vertex belonging to its own community maximizes. Experiments on both synthetic and several real-world networks (with known ground-truth community structures) reveal that EnCoD significantly outperforms nine state-of-the-art overlapping community detection algorithms. Finally, we show that EnCoD is generic enough to be applied to networks where the vertices are associated with explicit semantic features. To the best of our knowledge, EnCoD is the second ensemble-based overlapping community detection approach after MEDOC [1].

cs.SI↗

arXiv2018

Engineering chimera patterns in networks using heterogeneous delays

Symmetry breaking spatial patterns, referred to as chimera states, have recently been catapulted into the limelight due to their coexisting coherent and incoherent hybrid dynamics. Here, we present a method to engineer a chimera state by using an appropriate distribution of heterogeneous time delays on the edges of a network. The time delays in interactions, intrinsic to natural or artificial complex systems, are known to induce various modifications in spatiotemporal behaviors of the coupled dynamics on networks. Using a coupled chaotic map with the identical coupling environment, we demonstrate that control over the spatial location of the incoherent region of a chimera state in a network can be achieved by appropriately introducing time delays. This method allows for the engineering of tailor-made one cluster or multi-cluster chimera patterns. Furthermore, borrowing a measure of eigenvector localization from the spectral graph theory, we introduce a spatial inverse participation ratio, which provides a robust way for the identification of the chimera state. This report highlights the necessity to consider the heterogeneous time delays to develop applications for the chimera states in particular and understand coupled dynamical systems in general.

nlin.CD↗

arXiv2017

Non-identical multiplexing promotes chimera states

We present the emergence of chimeras, a state referring to coexistence of partly coherent, partly incoherent dynamics in networks of identical oscillators, in a multiplex network consisting of two non-identical layers which are interconnected. We demonstrate that the parameter range displaying the chimera state in the homogeneous first layer of the multiplex networks can be tuned by changing the link density or connection architecture of the same nodes in the second layer. We focus on the impact of the interconnected second layer on the enlargement or shrinking of the coupling regime for which chimeras are displayed in the homogeneous first layer. We find that a denser homogeneous second layer promotes chimera in a sparse first layer, where chimeras do not occur in isolation. Furthermore, while a dense connection density is required for the second layer if it is homogeneous, this is not true if the second layer is inhomogeneous. We demonstrate that a sparse inhomogeneous second layer which is common in real-world complex systems can promote chimera states in a sparse homogeneous first layer.

nlin.CD↗