INTERNATIONAL JOURNAL OF SCIENTIFIC DEVELOPMENT AND RESEARCH International Peer Reviewed & Refereed Journals, Open Access Journal ISSN Approved Journal No: 2455-2631 | Impact factor: 8.15 | ESTD Year: 2016
open access , Peer-reviewed, and Refereed Journals, Impact factor 8.15
Classifying User Age group using Deep Learning Technique
Authors Name:
Rajashree Anil Kale
, Prof Sarika B. Solanke
Unique Id:
IJSDR2001008
Published In:
Volume 5 Issue 1, January-2020
Abstract:
Online social networks have a lot of information. But often people do not provide personal information, such as age, gender and other demographic data, although the confidence analysis uses such information to develop useful applications in people's daily lives. But there is still a failure in this type of analysis, whether by the limited number of words contained in the word dictionary or because they do not consider the most diverse parameters Can influence feelings in sentences; Therefore, more reliable results will be obtained if considering user profile data and user writing style. This research shows that one of the most relevant parameters contained in the user profile is the age group, which shows that there is normal behavior among users of the same age group, especially when these users write about. With the same topic Detailed analysis with 7000 sentences has been conducted to determine which features are relevant, such as the use of punctuation, number of characters, sharing of media, other topics, and which ones can ignore the age group classification. Different learning machine algorithms have been tested for the classification of adolescent and adult groups and the Deep Convolutional Neural Network (DCNN) has the best performance with accuracy up to 0.95 in the validation test. must In addition, in order to verify the usefulness of the proposed model for age group classification, it is implemented in the Sentiment Metric (eSM) that has been improved. In performance audits, subjective tests are performed and eSM with the proposed model arrives. Mean Square root error and Pearson's correlation coefficient of 0.25 and 0.94, respectively, are more efficient than the eSM indicators when no age group information is specified.
Keywords:
machine learning, text analysis, artificial neural networks, deep network,Social network services, sentiment analysis,
Cite Article:
"Classifying User Age group using Deep Learning Technique", International Journal of Science & Engineering Development Research (www.ijsdr.org), ISSN:2455-2631, Vol.5, Issue 1, page no.46 - 50, January-2020, Available :http://www.ijsdr.org/papers/IJSDR2001008.pdf
Downloads:
000337074
Publication Details:
Published Paper ID: IJSDR2001008
Registration ID:191209
Published In: Volume 5 Issue 1, January-2020
DOI (Digital Object Identifier):
Page No: 46 - 50
Publisher: IJSDR | www.ijsdr.org
ISSN Number: 2455-2631
Facebook Twitter Instagram LinkedIn