Hassan Farooq
FAST
· 2020
·
16I-0302
Email
—
Phone
—
LinkedIn
—
GitHub
—
Academic
Program
—
CGPA
—
Year
2020
Education
—
Address
—
DOB
—
Career
Current role
—
Target role
—
Skills
Python, Colab, Keras, Flask, CNN, GRU, Deep Learning, NLP
Verbatim text
The exact text the LLM saw on the page (or the booklet text from the old import).
This is what powers semantic search.
Hate Speech Classification There are many sites and social media platforms who wants to restrict and block specific type of hate or discriminating content. Google, Facebook and twitter still remove hate content manually and it is always a serious problem in this evolving technology and that’s why this subject is catching attention from all over the world. In this research project, we carried out some experiments by keeping in view the previous research which used google word2vec embedding and other techniques like char n-gram method. We used best performing architecture CNN + GRU with flair embedding (GloVe, BERT, ELMo, Character Embeddings) which is new and its custom embeddings where we can use multiple embeddings at the same time. Our goal in this project was to improve the previous results with the state-of-the-art deep learning techniques and as a by product we made a web interface where we can classify hate speech into racism, sexism and non-hate by simply feeding a sentence into the model which is in the backend. \
Provenance
Source file: Graduate Directory FAST School of Computing 2020 (Final Complete) (1).pdfFrom job #23 page 197
Created: 1778226103