Abstract:Question-answering (QA) that comes naturally to humans is a critical component in seamless human-computer interaction. It has emerged as one of the most convenient and natural methods to interact with the web and is especially desirable in voice-controlled environments. Despite being one of the oldest research areas, the current QA system faces the critical challenge of handling multilingual queries. To build an Artificial Intelligent (AI) agent that can serve multilingual end users, a QA system is required to be language versatile and tailored to suit the multilingual environment. Recent advances in QA models have enabled surpassing human performance primarily due to the availability of a sizable amount of high-quality datasets. However, the majority of such annotated datasets are expensive to create and are only confined to the English language, making it challenging to acknowledge progress in foreign languages. Therefore, to measure a similar improvement in the multilingual QA system, it is necessary to invest in high-quality multilingual evaluation benchmarks. In this dissertation, we focus on advancing QA techniques for handling end-user queries in multilingual environments. This dissertation consists of two parts. In the first part, we explore multilingualism and a new dimension of multilingualism referred to as code-mixing. Second, we propose a technique to solve the task of multi-hop question generation by exploiting multiple documents. Experiments show our models achieve state-of-the-art performance on answer extraction, ranking, and generation tasks on multiple domains of MQA, VQA, and language generation. The proposed techniques are generic and can be widely used in various domains and languages to advance QA systems.

Learning to Translate for Multilingual Question Answering

Question Translation Training for Better Multilingual Reasoning

Learning to Answer Multilingual and Code-Mixed Questions

Cross-Lingual Transfer Learning for Question Answering

Bridging the Language Gap: Knowledge Injected Multilingual Question Answering

Survey of Multilingual Question Answering

Is Translation All You Need? A Study on Solving Multilingual Tasks with Large Language Models

The Power of Question Translation Training in Multilingual Reasoning: Broadened Scope and Deepened Insights

Statistical Machine Translation Improves Question Retrieval in Community Question Answering Via Matrix Factorization.

Translation Deserves Better: Analyzing Translation Artifacts in Cross-lingual Visual Question Answering

Evaluation Methodology for Large Language Models for Multilingual Document Question and Answer

Translation Language Model Enhancement For Community Question Retrieval Using User Adoption Answer

Modeling Bilingual Conversational Characteristics for Neural Chat Translation

Learning Inter-Related Statistical Query Translation Models for English-Chinese Bi-Directional CLIR

Knowledge-Based Question Answering As Machine Translation.

Towards Zero-Shot Multilingual Synthetic Question and Answer Generation for Cross-Lingual Reading Comprehension

Multilingual Multimodal Learning with Machine Translated Text

Cross-Lingual Training for Automatic Question Generation

Research on Question Classification via Bilingual Information

M2QA: Multi-domain Multilingual Question Answering

Addressing Community Question Answering in English and Arabic