| Autor: |
Manoj Kumar, Minakshi Tomer |
| Rok vydání: |
2020 |
| Předmět: |
|
| Zdroj: |
Advances in Information Communication Technology and Computing ISBN: 9789811554209 |
| DOI: |
10.1007/978-981-15-5421-6_28 |
| Popis: |
Question answering system is an area under natural language processing and information retrieval which automatically answers the questions generated by humans. This work represents an approach for building a system that generates answers for the question based on deep learning neural network which has the competence of processing the information present inside the dataset and enables the user to obtain an insight from the SQuAD dataset by inviting questions in natural language form. Key stages of this approach cover corpus pre-processing, question pre-processing, answer generation, deep neural network for answer extraction and keyword generation. The concept of keyword generation is a novel idea implemented to enable naive user of the system to apprehend the passage. The system is competent in interpreting the question, responding to the user’s query in natural language form along with generating the keywords. The performance was measured on SQuAD dataset using EM and F1 score. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
