scispace - formally typeset
Search or ask a question
Author

Utpal Kumar Sikdar

Bio: Utpal Kumar Sikdar is an academic researcher from Indian Institute of Technology Patna. The author has contributed to research in topics: Conditional random field & Named-entity recognition. The author has an hindex of 9, co-authored 32 publications receiving 777 citations. Previous affiliations of Utpal Kumar Sikdar include Norwegian University of Science and Technology.

Papers
More filters
Proceedings ArticleDOI
01 Aug 2017
TL;DR: A deep learning-based Twitter hate-speech text classification system that assigns each tweet to one of four predefined categories: racism, sexism, both (racism and sexism) and non-hate-speech.
Abstract: The paper introduces a deep learning-based Twitter hate-speech text classification system. The classifier assigns each tweet to one of four predefined categories: racism, sexism, both (racism and sexism) and non-hate-speech. Four Convolutional Neural Network models were trained on resp. character 4-grams, word vectors based on semantic information built using word2vec, randomly generated word vectors, and word vectors combined with character n-grams. The feature set was down-sized in the networks by max-pooling, and a softmax function used to classify tweets. Tested by 10-fold cross-validation, the model based on word2vec embeddings performed best, with higher precision than recall, and a 78.3% F-score.

419 citations

Journal ArticleDOI
TL;DR: The CHEMDNER corpus is presented, a collection of 10,000 PubMed abstracts that contain a total of 84,355 chemical entity mentions labeled manually by expert chemistry literature curators, following annotation guidelines specifically defined for this task.
Abstract: The automatic extraction of chemical information from text requires the recognition of chemical entity mentions as one of its key steps. When developing supervised named entity recognition (NER) systems, the availability of a large, manually annotated text corpus is desirable. Furthermore, large corpora permit the robust evaluation and comparison of different approaches that detect chemicals in documents. We present the CHEMDNER corpus, a collection of 10,000 PubMed abstracts that contain a total of 84,355 chemical entity mentions labeled manually by expert chemistry literature curators, following annotation guidelines specifically defined for this task. The abstracts of the CHEMDNER corpus were selected to be representative for all major chemical disciplines. Each of the chemical entity mentions was manually labeled according to its structure-associated chemical entity mention (SACEM) class: abbreviation, family, formula, identifier, multiple, systematic and trivial. The difficulty and consistency of tagging chemicals in text was measured using an agreement study between annotators, obtaining a percentage agreement of 91. For a subset of the CHEMDNER corpus (the test set of 3,000 abstracts) we provide not only the Gold Standard manual annotations, but also mentions automatically detected by the 26 teams that participated in the BioCreative IV CHEMDNER chemical mention recognition task. In addition, we release the CHEMDNER silver standard corpus of automatically extracted mentions from 17,000 randomly selected PubMed abstracts. A version of the CHEMDNER corpus in the BioC format has been generated as well. We propose a standard for required minimum information about entity annotations for the construction of domain specific corpora on chemical and drug entities. The CHEMDNER corpus and annotation guidelines are available at: http://www.biocreative.org/resources/biocreative-iv/chemdner-corpus/

368 citations

Proceedings Article
01 Dec 2012
TL;DR: The proposed differential evolution (DE) based feature selection and classifier ensemble methods that can be applied to any classification problem and scope of the development of language independent NER systems in a resource-poor environment are evaluated.
Abstract: In this paper, we propose a differential evolution (DE) based two-stage evolutionary approach for named entity recognition (NER). The first stage concerns with the problem of relevant feature selection for NER within the frameworks of two popular machine learning algorithms, namely Conditional Random Field (CRF) and Support Vector Machine (SVM). The solutions of the final best population provides different diverse set of classifiers; some are effective with respect to recall whereas some are effective with respect to precision. In the second stage we propose a novel technique for classifier ensemble for combining these classifiers. The approach is very general and can be applied for any classification problem. Currently we evaluate the proposed algorithm for NER in three popular Indian languages, namely Bengali, Hindi and Telugu. In order to maintain the domain-independence property the features are selected and developed mostly without using any deep domain knowledge and/or language dependent resources. Experimental results show that the proposed two stage technique attains the final F-measure values of 88.89%, 88.09% and 76.63% for Bengali, Hindi and Telugu, respectively. The key contributions of this work are two-fold, viz. (i). proposal of differential evolution (DE) based feature selection and classifier ensemble methods that can be applied to any classification problem; and (ii). scope of the development of language independent NER systems in a resource-poor

30 citations

Journal ArticleDOI
01 Dec 2015
TL;DR: The key contribution of this work is the development of MODE-based generalized feature selection and ensemble learning techniques with the aim of extracting entities from the biomedical texts of several domains.
Abstract: In this paper, we propose a multiobjective differential evolution (MODE)-based feature selection and ensemble learning approaches for entity extraction in biomedical texts. The first step of the algorithm concerns with the problem of automatic feature selection in a machine learning framework, namely conditional random field. The final Pareto optimal front which is obtained as an output of the feature selection module contains a set of solutions, each of which represents a particular feature representation. In the second step of our algorithm, we combine a subset of these classifiers using a MODE-based ensemble technique. Our experiments on three benchmark datasets namely GENIA, GENETAG and AIMed show the F-measure values of 76.75, 94.15 and 91.91 %, respectively. Comparisons with the existing systems show that our proposed algorithm achieves the performance levels which are at par with the state of the art. These results also exhibit that our method is general in nature and because of this it performs well across the several domain of datasets. The key contribution of this work is the development of MODE-based generalized feature selection and ensemble learning techniques with the aim of extracting entities from the biomedical texts of several domains.

27 citations

Journal ArticleDOI
01 Aug 2015
TL;DR: A differential evolution (DE)-based feature selection technique is developed for anaphora resolution in a resource-poor language, namely Bengali and a number of models for mention detection based on machine learning and heuristics are developed.
Abstract: In this paper a differential evolution (DE)-based feature selection technique is developed for anaphora resolution in a resource-poor language, namely Bengali. We discuss the issues of adapting a state-of-the-art English anaphora resolution system for a resource-poor language like Bengali. Performance of any anaphoric resolver greatly depends on the quality of a high accurate mention detector and the use of appropriate features for anaphora resolution. We develop a number of models for mention detection based on machine learning and heuristics. In anaphora resolution there is no globally accepted metric for measuring the performance, and each of them such as MUC, $$\hbox {B}^{3}$$B3, CEAF, Blanc exhibit significantly different behaviors. Our proposed feature selection technique determines the near-optimal feature set by optimizing each of these evaluation metrics. Experiments show how a language-dependent system (designed primarily for English) can attain reasonably good performance level when re-trained and tested on a new language with a proper subset of features. Evaluation results yield the F-measure values of 66.70, 59.47, 51.56, 33.08 and 72.75 % for MUC, B 3, CEAFM, CEAFE and BLANC, respectively.

27 citations


Cited by
More filters
Journal ArticleDOI
TL;DR: Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis.
Abstract: Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

13,246 citations

01 Jan 2002

9,314 citations

Journal ArticleDOI
TL;DR: This article proposed BioBERT (Bidirectional Encoder Representations from Transformers for Biomedical Text Mining), which is a domain-specific language representation model pre-trained on large-scale biomedical corpora.
Abstract: Motivation Biomedical text mining is becoming increasingly important as the number of biomedical documents rapidly grows. With the progress in natural language processing (NLP), extracting valuable information from biomedical literature has gained popularity among researchers, and deep learning has boosted the development of effective biomedical text mining models. However, directly applying the advancements in NLP to biomedical text mining often yields unsatisfactory results due to a word distribution shift from general domain corpora to biomedical corpora. In this article, we investigate how the recently introduced pre-trained language model BERT can be adapted for biomedical corpora. Results We introduce BioBERT (Bidirectional Encoder Representations from Transformers for Biomedical Text Mining), which is a domain-specific language representation model pre-trained on large-scale biomedical corpora. With almost the same architecture across tasks, BioBERT largely outperforms BERT and previous state-of-the-art models in a variety of biomedical text mining tasks when pre-trained on biomedical corpora. While BERT obtains performance comparable to that of previous state-of-the-art models, BioBERT significantly outperforms them on the following three representative biomedical text mining tasks: biomedical named entity recognition (0.62% F1 score improvement), biomedical relation extraction (2.80% F1 score improvement) and biomedical question answering (12.24% MRR improvement). Our analysis results show that pre-training BERT on biomedical corpora helps it to understand complex biomedical texts. Availability and implementation We make the pre-trained weights of BioBERT freely available at https://github.com/naver/biobert-pretrained, and the source code for fine-tuning BioBERT available at https://github.com/dmis-lab/biobert.

2,680 citations

Journal ArticleDOI
01 Jan 2016-Database
TL;DR: The BC5CDR corpus was successfully used for the BioCreative V challenge tasks and should serve as a valuable resource for the text-mining research community.
Abstract: Community-run, formal evaluations and manually annotated text corpora are critically important for advancing biomedical text-mining research. Recently in BioCreative V, a new challenge was organized for the tasks of disease named entity recognition (DNER) and chemical-induced disease (CID) relation extraction. Given the nature of both tasks, a test collection is required to contain both disease/chemical annotations and relation annotations in the same set of articles. Despite previous efforts in biomedical corpus construction, none was found to be sufficient for the task. Thus, we developed our own corpus called BC5CDR during the challenge by inviting a team of Medical Subject Headings (MeSH) indexers for disease/chemical entity annotation and Comparative Toxicogenomics Database (CTD) curators for CID relation annotation. To ensure high annotation quality and productivity, detailed annotation guidelines and automatic annotation tools were provided. The resulting BC5CDR corpus consists of 1500 PubMed articles with 4409 annotated chemicals, 5818 diseases and 3116 chemical-disease interactions. Each entity annotation includes both the mention text spans and normalized concept identifiers, using MeSH as the controlled vocabulary. To ensure accuracy, the entities were first captured independently by two annotators followed by a consensus annotation: The average inter-annotator agreement (IAA) scores were 87.49% and 96.05% for the disease and chemicals, respectively, in the test set according to the Jaccard similarity coefficient. Our corpus was successfully used for the BioCreative V challenge tasks and should serve as a valuable resource for the text-mining research community.Database URL: http://www.biocreative.org/tasks/biocreative-v/track-3-cdr/.

605 citations