Download PDFOpen PDF in browser
Switch back to the title and the abstract in Turkish

Multi-Label Text Analysis with GRU and CNN Based Hybrid Deep Learning Model

EasyChair Preprint no. 8449

6 pagesDate: July 10, 2022


Written shares shared on social networks such as Twitter, Facebook, Instagram, WhatsApp and Wikipedia are increasing day by day. Data content on social networks is increasing without any control. It is important to automatically analyze the meaningful, positive, and negative data content among this increasing written social sharing data. For this purpose, a text analysis study was carried out to analyze the content of written shares belonging to the Wikipedia social sharing system. Although it is possible to preprocess and digitize data with text analysis in content analysis, deep learning models are needed to obtain the feature map. Natural language processing methods were used to perform text analysis. Multi-label text classification according to the content of the analyzed texts was carried out with a hybrid model based on Gated Recurrent Unit (GRU) and Convolutional Neural Network (CNN). Natural language processing techniques are preprocessed after taking Wikipedia user comments as input. The preprocessed text data are digitized with the GloVe embedding layer. Digitized texts are processed by giving input to multi-layered deep learning architecture. The classification of thousands of Wikipedia comments with deep learning, which is one of the active subfields of machine learning, was carried out with the specified steps. The performance metrics obtained in the experimental studies carried out with the proposed method are presented.

Keyphrases: CNN, derin öğrenme, GRU, Wikipedia, Çok etiketli metin analizi

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
  author = {Halit Çetiner},
  title = {Multi-Label Text Analysis with GRU and CNN Based Hybrid Deep Learning Model},
  howpublished = {EasyChair Preprint no. 8449},

  year = {EasyChair, 2022}}
Download PDFOpen PDF in browser