Skip to content

kianomoomi/Important-News-Detection

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 

Repository files navigation

Important News Detection

Task Description

In everyday life, news may be published that can have a significant impact on people’s lives. Reading all the news and finding the most important and impactful news is not feasible for everyone and is a challenging task. Therefore, having a model that can distinguish important news from non-important news can be highly beneficial.

Proposed Method

In this research, we use one of these language models to categorize important news and also examine the impact of the prompt length given to the model. In fact, our ultimate goal is to examine the effect of the text length or prompt given to the large language model in detecting important news.
For this purpose, we generate different prompts and provide them to the large language model, and compare the results. The prompt consists of a general task description at the beginning, followed by a description of some generally important subjects. Then, (k) similar news items to the target news are provided, and finally, the target news is included in the prompt. The distinguishing factors of the prompts are: the type of news whose importance needs to be detected and the number of similar news items included in each prompt. This in-context learning approach can benefit the model in learning the concept.
For each news item under study, three different cases are presented to the model: the news title, the full text, and a summary of it. By analyzing the results of the various combinations of these factors, we can evaluate the impact of the text length or prompt on the model’s ability to detect important news.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published