Job Prediction: From Deep Neural Network Models to Applications

Abstract

Determining the job is suitable for a student or a person looking for work based on their job descriptions such as knowledge and skills that are difficult, as well as how employers must find ways to choose the candidates that match the job they require. In this paper, we focus on studying the job prediction using different deep neural network models including TextCNN, Bi-GRU-LSTM-CNN, and Bi-GRU-CNN with various pre-trained word embeddings on the IT job dataset. In addition, we proposed a simple and effective ensemble model combining different deep neural network models. Our experimental results illustrated that our proposed ensemble model achieved the highest result with an F1-score of 72.71%. Moreover, we analyze these experimental results to have insights about this problem to find better solutions in the future.

UIT-HSE at WNUT-2020 Task 2: Exploiting CT-BERT for Identifying COVID-19 Information on the Twitter Social Network

“…Fine-tuning pre-trained language models (Dodge et al, 2020) also have proved the effectiveness of exploring solutions for new tasks. In addition, ensemble approaches were also very effective in previous studies (Araque et al, 2017;Nguyen et al, 2019;Van Huynh et al, 2020). Inspired by these research works, we propose our simple but effective ensemble strategy using different models based on CT-BERT with different fine-tunings, achieving the F1-score of 90.94% on the final test set.…”

BANANA at WNUT-2020 Task 2: Identifying COVID-19 Information on Twitter by Combining Deep Learning and Transfer Learning Models

“…As the success of the ensemble models of previous tasks (Van Huynh et al, 2020;, we propose a simple yet effective ensemble approach with the majority voting between the outputs of four different models, including Bi-GRU-CNN, BERT, RoBERTa, and XLNet for classifying whether a tweet contains information about COVID-19 or not. 5 Experiment…”

Title2Vec: a contextual job title embedding for occupational named entity recognition and other applications

“…Table 1 shows our literature review of the related publications that utilizes these types of occupationalrelated dataset. Out of these 23 datasets, only two been made publicly available, which are [1] by James et al [25] by Van Huynh et al and [10] that is described in this paper. The former dataset by James et al [1] contains the affiliations that physics researchers belong to but does not include their job titles or positions, whereas our dataset comprises the job titles of general workers across the broader industry.…”

“…Table 1 A survey of related works that uses and/or provide occupational-related datasets Apart from two datasets of comprising publications and authors [1] and job titles and descriptions [25], there are no publicly available occupational-related dataset from our survey. The first dataset [1] contains publications and authors from the American Physics Society (APS) but only describes the names and affiliations of physics scientists without their job title or appointments, while the second dataset [25] contains the job title and job description from a job portal but only pertaining to IT-related jobs. Our proposed dataset, IPOD is bolded and in the first row…”