FinABSA Open-Source Financial Sentiment Analysis Model - Accurately Analyze the Sentiment Tendencies of Sentences from Multiple Aspects

Finabsa

Developed by amphora

FinABSA is a model trained on the T5-Large architecture for Aspect-Based Sentiment Analysis (ABSA) tasks. It can parse sentences containing multiple aspects. By replacing the target aspect with the [TGT] token, the model can focus on sentiment prediction for that specific aspect.

Large Language Model

Transformers

EnglishOpen Source License:Apache-2.0 #Financial sentiment analysis #Aspect-level sentiment recognition #T5 architecture optimization

Downloads 146

Release Time : 11/2/2022

Model Overview

FinABSA is a sentiment analysis model for the financial domain, specifically designed to analyze sentiment tendencies toward specific aspects in text. Unlike traditional sentiment analysis models, it can perform independent sentiment analysis for different aspects within a sentence.

Model Features

Aspect-Based Sentiment Analysis

Capable of identifying and analyzing sentiment tendencies toward specific aspects in a sentence, rather than making a single sentiment judgment for the entire sentence.

Financial Domain Optimization

Trained using the financial domain dataset SEntFiN 1.0, making it particularly suitable for analyzing sentiment tendencies in financial texts.

Multi-Target Analysis

Can simultaneously analyze sentiment tendencies for multiple different aspects in a sentence by simply replacing the targets with [TGT] tokens.

Model Capabilities

Financial text sentiment analysis

Specific aspect sentiment recognition

Multi-target sentiment analysis

Use Cases

Financial Analysis

Stock Market Sentiment Analysis

Analyze sentiment tendencies toward specific company stocks in news reports

Accurately identifies sentiment tendencies for specific stocks in complex sentences

Corporate Earnings Report Sentiment Analysis

Extract emotional reactions to specific aspects of a company from earnings news

Can distinguish emotional attitudes toward different business units or departments of a company

Market Research

Product Review Analysis

Analyze sentiment tendencies toward specific product features in consumer reviews

Can precisely identify product strengths and weaknesses

🚀 FinABSA: Financial Aspect-Based Sentiment Analysis Model

FinABSA is a T5-Large model designed for aspect-based sentiment analysis in the financial domain, offering more nuanced sentiment insights.

🚀 Quick Start

You can use this model directly using the AutoModelForSeq2SeqLM class.

>>> from transformers import AutoModelForSeq2SeqLM, AutoTokenizer

>>> tokenizer = AutoTokenizer.from_pretrained("amphora/FinABSA")
>>> model = AutoModelForSeq2SeqLM.from_pretrained("amphora/FinABSA")

>>> input_str = "[TGT] stocks dropped 42% while Samsung rallied."
>>> input = tokenizer(input_str, return_tensors='pt')
>>> output = model.generate(**input, max_length=20)
>>> print(output)
The sentiment for [TGT] in the given sentence is NEGATIVE.

>>> input_str = "Tesla stocks dropped 42% while [TGT] rallied."
>>> input = tokenizer(input_str, return_tensors='pt')
>>> output = model.generate(**input, max_length=20)
>>> print(output)
The sentiment for [TGT] in the given sentence is POSITIVE.

✨ Features

Aspect-Based Analysis: Unlike traditional sentiment analysis models which predict a single sentiment label for each sentence, FinABSA has been trained to disambiguate sentences containing multiple aspects. By replacing the target aspect with a [TGT] token the model predicts the sentiment concentrating to the aspect.
Trained on Financial Data: FinABSA is trained for Aspect-Based Sentiment Analysis(ABSA) tasks using SEntFiN 1.0.

📚 Documentation

Model Description

FinABSA is a T5-Large model trained for Aspect-Based Sentiment Analysis(ABSA) tasks using SEntFiN 1.0. Unlike traditional sentiment analysis models which predict a single sentiment label for each sentence, FinABSA has been trained to disambiguate sentences containing multiple aspects. By replacing the target aspect with a [TGT] token the model predicts the sentiment concentrating to the aspect. GitHub Repo

Evaluation Results

Using a test split arbitarly extracted from SEntFiN 1.0 the model scores an average accuracy of 87%.

📄 License

This project is licensed under the Apache-2.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご