Resume Classification Using Bidirectional LSTM and Attention Mechanism

Bera, Swayami; Ghosh, Bikramaditya; Vanusha, D.

doi:10.1007/978-981-16-7952-0_22

Cited by 4 publications

(3 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Substituting the random forest classifier with logistic regression resulted in a drop in accuracy by 1-2%. We also compared our results with that of GloVe and Word2Vec word embeddings using BiLSTM with and without attention mechanism [6,12,7,25,37]; the attention mechanism has known to boost the accuracies of sequence learning models in the past.…”

Section: Discussion On Resultsmentioning

confidence: 99%

“…Several researchers have explored word embeddings for resume classification. Some notable and distinctive works in this direction are: -(a) GloVe word embeddings with convolutional neural networks [24] (b) Word2Vec word embeddings with Bidirectional LSTM with attention [25] (c) GloVe word embeddings with deep neural networks [26] (d) GloVe word embeddings with graph neural networks [27]. In [28], Zu and Wang claimed that the combination of Bidirectional LSTM with convolutional neural network and conditional random fields is the best classifier to classify word embedding sequences extracted from text blocks in a resume.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Uniqueness meets Semantics: A Novel Semantically Meaningful Bag-of-Words Approach for Matching Resumes to Job Profiles

Susan,

Sharma,

Choudhary

2024

View full text Add to dashboard Cite

In an increasingly competitive world, automated screening of resumes of candidates is the need of the hour given the large numbers of such resumes in career portals on the World Wide Web. Resume classification is a subset of the document classification problem in which the keywords extracted from the resume play a significant role in determining the job profile. In this paper, we explore the novel combination of uniqueness in terms of the number of occurrences of a keyword in a resume class as compared to the other resume classes, and the concept of semantics by representing the filtered keywords using word embeddings that can be used to find semantic similarities between resume documents. The principle of maximum entropy partitioning is used to find the keywords unique to a particular class. The aim is to use semantic representations of only those keywords that occur more frequently in one class more than in any other class; these are then passed as input to a Bidirectional long short-term memory (LSTM) for classification. Our experiments on a benchmark dataset proves that the proposed approach outperforms the state of the art in text classification by a significant margin proving the efficacy of our approach.

show abstract

Section: Discussion On Resultsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Uniqueness meets Semantics: A Novel Semantically Meaningful Bag-of-Words Approach for Matching Resumes to Job Profiles

Susan,

Sharma,

Choudhary

2024

View full text Add to dashboard Cite

show abstract

“…It uses convolutional neural networks (CNN) and recurrent neural networks (RNN) to extract features from CVs and classify them into speci c categories. [7] and [8] has explored different approaches and techniques to get the most out of this model. Researchers have proposed model architectures speci cally tailored to BERT, using additional layers for classi cation, or in combination with other machine learning methods such as convolutional neural networks (CNN) or recurrent neural networks (RNN).…”

Section: Related Workmentioning

confidence: 99%

CV classification in a big data environment using BERT and alignment learning

CHAFI,

KABIL,

KAMOUSS

2023

Preprint

View full text Add to dashboard Cite

Today, text classification has always been a crucial discipline in the field of text data processing. However, with the advent of Big Data, text classification has reached new heights in terms of the volume of data to be processed and the complexity of the tasks. This task is of great importance in many fields, such as spam detection, sentiment analysis, document categorization, content recommendation and many more. In this work, we focus specifically on resume classification, an area that presents unique challenges due to diverse formats, ambiguous language, and variations in applicants' work experiences. We propose an innovative approach for distributed classification of CVs using contextual alignment techniques with the job offer and the pre-trained language model BERT (Bidirectional Encoder Representations from Transformers) which constitutes a major advance in this field , and using RNN (Recurrent Neural Network) to solve text classification problems. We use a distributed processing approach to leverage the parallel computing power of Spark, enabling large volumes of data to be processed efficiently. The main objective of our study is to improve the relevance of CV classification by leveraging the pre-trained language models and distributed processing power of Spark.

show abstract