Building domain specific lexicon based on TikTok comment dataset

12/16/2020
by   Hao Jiaxiang, et al.
0

In the sentiment analysis task, predicting the sentiment tendency of a sentence is an important branch. Previous research focused more on sentiment analysis in English, for example, analyzing the sentiment tendency of sentences based on Valence, Arousal, Dominance of sentences. the emotional tendency is different between the two languages. For example, the sentence order between Chinese and English may present different emotions. This paper tried a method that builds a domain-specific lexicon. In this way, the model can classify Chinese words with emotional tendency. In this approach, based on the [13], an ultra-dense space embedding table is trained through word embedding of Chinese TikTok review and emotional lexicon sources(seed words). The result of the model is a domain-specific lexicon, which presents the emotional tendency of words. I collected Chinese TikTok comments as training data. By comparing The training results with the PCA method to evaluate the performance of the model in Chinese sentiment classification, the results show that the model has done well in Chinese. The source code has released on github:https://github.com/h2222/douyin_comment_dataset

READ FULL TEXT

page 1

page 2

page 3

page 4

research
10/18/2022

Sentiment-Aware Word and Sentence Level Pre-training for Sentiment Analysis

Most existing pre-trained language representation models (PLMs) are sub-...
research
08/01/2021

Transformer-Encoder-GRU (T-E-GRU) for Chinese Sentiment Analysis on Chinese Comment Text

Chinese sentiment analysis (CSA) has always been one of the challenges i...
research
08/10/2017

Radical-level Ideograph Encoder for RNN-based Sentiment Analysis of Chinese and Japanese

The character vocabulary can be very large in non-alphabetic languages s...
research
05/01/2019

Emotional Contribution Analysis of Online Reviews

In response to the constant increase in population and tourism worldwide...
research
06/09/2016

Inducing Domain-Specific Sentiment Lexicons from Unlabeled Corpora

A word's sentiment depends on the domain in which it is used. Computatio...
research
09/18/2023

The ParlaSent multilingual training dataset for sentiment identification in parliamentary proceedings

Sentiments inherently drive politics. How we receive and process informa...

Please sign up or login with your details

Forgot password? Click here to reset