Question Answering by Bert
Suman Karanjit, Minnesota State University MoorheadFollow
2021 12:00 AM
There are many machine learning algorithms that tries to solve the question answering using the natural language. In old day, the bag of words was popular amongst others which tries to answer the question that are pre-defined by the developers. Using this method, developers have to spend a lot of time writing the questions and answer for the particular questions. This method was very useful for the chat bots but was not able to answer the questions for the huge database. Modern Natural Language Processor are dominated by the models called transformers. Using these transformers library, the Bert QA model was introduced by HuggingFace which reads through the text context provided by the user and tries to answer the questions related to that text context. This model has been promising in answering the complex question from a large document. For example, if one company have a report regarding their financial years that been fed through the model, user can just ask question regarding the certain year or the profits they made for particular year. And without scrolling through the documents the answer can be found with the matter of seconds.
Since April 23, 2021
Artificial Intelligence and Robotics Commons,
Data Science Commons
To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.
NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.
My Account |