KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#developer-tools #typescript #claude-code #codex #dsh #dsh-plugin #ai-agent #ai-agents #deepseek-harness #agent-skills #macos #cli

multi-domain-sentiment

★ 30 Open GitHub ↗

Code for the paper Learning Domain-specific Representations for Multi-Domain Sentiment Classification

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

multi-domain-sentiment
======
A framework for multi-domain sentiment analysis by learning domain-specific representations of input sentences using neural network.

Prerequisite
======
1. Tensorflow 
2. Google News Embeddings (https://code.google.com/archive/p/word2vec/) (rename it to 'vectors.gz' and put it under the main folder)
3. Gensim

Data Preparation
======
1. Download datasets (e.g. laptops). We assume the datasets are preprocessed into the following format:

The unit does everything it promises . I 've only used it once so far , but i 'm happy with it ||| 1

2. Randomly split each dataset into training (e.g. laptops/trn), development (e.g. laptops/dev) and testing datasets (e.g. laptops/tst). Put all datasets into a folder named 'dataset'. Thus, the directory structure looks like dataset/laptops/trn.

Preprocessing and Run the Demo
======

1. Run `python preprocessing.py`. This program will iterate through the 'dataset' folder and generate files like dictionaries, embeddings and transformed datasets.

2. Run `python multi_view_domain_embedding_memory_adversarial.py dataset_name1 dataset_name2 ...` for running the algorithm.

multi-domain-sentiment

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README