Please use this identifier to cite or link to this item:
http://dspace.dtu.ac.in:8080/jspui/handle/repository/19084
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | PATIDAR, ASHISH | - |
dc.date.accessioned | 2022-06-07T06:04:44Z | - |
dc.date.available | 2022-06-07T06:04:44Z | - |
dc.date.issued | 2021-06 | - |
dc.identifier.uri | http://dspace.dtu.ac.in:8080/jspui/handle/repository/19084 | - |
dc.description.abstract | Sarcasm is one of the sentiments which is being used to communicate a negative opinion utilizing positive words. The world is full of social media and many kinds of the web-based portal and this media stores a huge amount of textual data which contains sentiments, and sarcasm is one of the sentiments which is being used nowadays in many of this platform, using sarcasm someone can communicate their negative words in a positive way which is we can call a sarcastic way of communication. In opinion mining, the field of natural language processing detection of sarcasm from a given data is an important task. It is a binary classification task for which model proposed a system which classifies whether a given set of word is sarcastic or not-sarcastic. In this research work, we proposed the work based on the Stacked Bi-Directional Long Short-Term Memory (Stk BLSTM) network which enhances the overall result in terms of performance matrix. | en_US |
dc.language.iso | en | en_US |
dc.relation.ispartofseries | TD-5630; | - |
dc.subject | SARCASM DETECTION | en_US |
dc.subject | NATURAL LANGUAGE PROCESSING | en_US |
dc.subject | STACKED LSTM | en_US |
dc.subject | MACHINE LEARNING | en_US |
dc.subject | SOCIAL DATA | en_US |
dc.title | SARACASM DETECTION USING STACKED BI-DIRECTIONAL LSTM MODE | en_US |
dc.type | Thesis | en_US |
Appears in Collections: | M.E./M.Tech. Computer Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
ASHISH PATIDAR M.Tech..pdf | 1.09 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.