Searcharxiv⌕ Search

arXiv subjects

Arindam Mitra

Publications and source records attributed to Arindam Mitra.

At least 55 records · Page 3Linked to original sources
arXiv2019

Careful Selection of Knowledge to solve Open Book Question Answering

Open book question answering is a type of natural language based QA (NLQA) where questions are expected to be answered with respect to a given set of open book facts, and common knowledge about a topic. Recently a challenge involving such QA, OpenBookQA, has been proposed. Unlike most other NLQA tasks that focus on linguistic understanding, OpenBookQA requires deeper reasoning involving linguistic understanding as well as reasoning with common knowledge. In this paper we address QA with respect to the OpenBookQA dataset and combine state of the art language models with abductive information retrieval (IR), information gain based re-ranking, passage selection and weighted scoring to achieve 72.0% accuracy, an 11.6% improvement over the current state of the art.

cs.CL↗
arXiv2019

Declarative Question Answering over Knowledge Bases containing Natural Language Text with Answer Set Programming

While in recent years machine learning (ML) based approaches have been the popular approach in developing end-to-end question answering systems, such systems often struggle when additional knowledge is needed to correctly answer the questions. Proposed alternatives involve translating the question and the natural language text to a logical representation and then use logical reasoning. However, this alternative falters when the size of the text gets bigger. To address this we propose an approach that does logical reasoning over premises written in natural language text. The proposed method uses recent features of Answer Set Programming (ASP) to call external NLP modules (which may be based on ML) which perform simple textual entailment. To test our approach we develop a corpus based on the life cycle questions and showed that Our system achieves up to $18\%$ performance gain when compared to standard MCQ solvers.

cs.AI↗
arXiv2019

Understanding Roles and Entities: Datasets and Models for Natural Language Inference

We present two new datasets and a novel attention mechanism for Natural Language Inference (NLI). Existing neural NLI models, even though when trained on existing large datasets, do not capture the notion of entity and role well and often end up making mistakes such as "Peter signed a deal" can be inferred from "John signed a deal". The two datasets have been developed to mitigate such issues and make the systems better at understanding the notion of "entities" and "roles". After training the existing architectures on the new dataset we observe that the existing architectures does not perform well on one of the new benchmark. We then propose a modification to the "word-to-word" attention function which has been uniformly reused across several popular NLI architectures. The resulting architectures perform as well as their unmodified counterparts on the existing benchmarks and perform significantly well on the new benchmark for "roles" and "entities".

cs.CL↗
arXiv2018

Incremental and Iterative Learning of Answer Set Programs from Mutually Distinct Examples

Over the years the Artificial Intelligence (AI) community has produced several datasets which have given the machine learning algorithms the opportunity to learn various skills across various domains. However, a subclass of these machine learning algorithms that aimed at learning logic programs, namely the Inductive Logic Programming algorithms, have often failed at the task due to the vastness of these datasets. This has impacted the usability of knowledge representation and reasoning techniques in the development of AI systems. In this research, we try to address this scalability issue for the algorithms that learn answer set programs. We present a sound and complete algorithm which takes the input in a slightly different manner and performs an efficient and more user controlled search for a solution. We show via experiments that our algorithm can learn from two popular datasets from machine learning community, namely bAbl (a question answering dataset) and MNIST (a dataset for handwritten digit recognition), which to the best of our knowledge was not previously possible. The system is publicly available at https://goo.gl/KdWAcV. This paper is under consideration for acceptance in TPLP.

cs.AI↗
arXiv2009

Human being is a living random number generator

General wisdom is, mathematical operation is needed to generate number by numbers. It is pointed out that without any mathematical operation true random numbers can be generated by numbers through algorithmic process. It implies that human brain itself is a living true random number generator. Human brain can meet the enormous human demand of true random numbers.

cs.DS↗
arXiv2009

Quantum theory can be collectively verified

No theory of physics has been collectively scientifically verified in an experiment so far. It is pointed out that probabilistic structure of quantum theory can be collectively scientifically verified in an experiment. It is also argued that experimentalist point of view quantum theory is a complete theory.

cs.OH↗
arXiv2009

Quantum mechanics can provide unbiased result

Getting an unbiased result is a remarkably long standing problem of collective observation/measurement. It is pointed out that quantum coin tossing can generate unbiased result defeating dishonesty.

cs.CR↗
arXiv2008

An unbreakable cryptosystem

The remarkably long-standing problem of cryptography is to generate completely secure key. It is widely believed that the task cannot be achieved within classical cryptography. However, there is no proof in support of this belief. We present an incredibly simple classical cryptosystem which can generate completely secure key.

cs.CR↗
arXiv2008

No-cloning principal can alone provide security

Existing quantum key distribution schemes need the support of classical authentication scheme to ensure security. This is a conceptual drawback of quantum cryptography. It is pointed out that quantum cryptosystem does not need any support of classical cryptosystem to ensure security. No-cloning principal can alone provide security in communication. Even no-cloning principle itself can help to authenticate each bit of information. It implies that quantum password need not to be a secret password.

cs.IT↗
arXiv2008

A particle can carry more than one bit of information

It is believed that a particle cannot carry more than one bit of information. It is pointed out that particle or single-particle quantum state can carry more than one bit of information. It implies that minimum energy cost of transmitting a bit will be less than the accepted limit KTlog2.

cs.IT↗
arXiv2008

Quantum communication is possible with pure state

It is believed that quantum communication is not possible with a pure ensemble of states because quantum entropy of pure state is zero. This is indeed possible due to geometric consequence of entanglement.

cs.IT↗
arXiv2007

An unbreakable cryptosystem for common people

It has been found that an algorithm can generate true random numbers on classical computer. The algorithm can be used to generate unbreakable message PIN (personal identification number) and password.

cs.CR↗
arXiv2005

Entangled messages

It is sometimes necessary to send copies of the same email to different parties, but it is impossible to ensure that if one party reads the message the other parties will bound to read it. We propose an entanglement based scheme where if one party reads the message the other party will bound to read it simultaneously.

cs.CR↗
arXiv2001

Space-Time Encoding

It is widely believed that one signal can carry maximum one bit of information, and minimum KTlog2 energy is required to transmit or to erase a bit. Here it is shown four bits of information can be transmitted by one signal in four dimensional space-time. It implies space-time cannot be ignored to determine the minimal energy cost of information processing.

physics.gen-ph↗
arXiv2001

Complete Quantum Communication with Security

The long-standing problem of quantum information processing is to remove the classical channel from quantum communication. Introducing a new information processing technique, it is discussed that both insecure and secure quantum communications are possible without the requirement of classical channel.

quant-ph↗
arXiv2001

Unconditionally secure quantum bit commitment is simply possible

Mayers, Lo and Chau proved unconditionally secure quantum bit commitment is impossible. It is shown that their proof is valid only for a particular model of quantum bit commitment encoding, in general it does not hold good. A different model of unconditionally secure quantum bit commitment - both entanglement and disentanglement-based - is presented. Even cheating can be legally proved with some legal evidences. Unconditionally secure quantum bit commitment can be established on the top of unconditionally secure quantum coin tossing, which is also claimed to be two-way impossible.

physics.gen-ph↗