Machine Learning Researcher - Search Team at ByteDance

Role Overview

Join the Search Team at ByteDance, dedicated to innovating search algorithms and systems for flagship products such as Douyin, Toutiao, Xigua Video, and additional services including E-commerce and Local Services. This role involves leveraging advanced machine learning techniques to enhance end-to-end search solutions, optimize distributed and machine learning architectures, and advance the frontiers of search technology.

Key Responsibilities

• Investigate state-of-the-art NLP technologies ranging from foundational tasks like word segmentation and Named Entity Recognition (NER) to complex implementations including text and multimodal pre-training, query analysis, and relevance modeling using deep learning.
• Develop cross-modal semantic matching approaches combining Computer Vision and Natural Language Processing to improve multimodal video search capabilities.
• Apply large-scale streaming machine learning techniques to personalize and refine search recommendation systems.
• Conduct architecture and system innovation for handling data at massive scales, overseeing offline computation, distributed system optimization, and designing low-latency, high-throughput online services.
• Advance recommendation algorithms in search through large-scale machine learning to deliver industry-leading personalized results.

Specific Research Targets

• Integrate large-scale model techniques with ranking algorithms to enhance personalization accuracy and user satisfaction.
• Explore generative retrieval algorithms tailored for ultra-large candidate sets containing tens to hundreds of billions of entries.
• Leverage large language models to improve search effectiveness for complex, ambiguous queries.
• Architect scalable, efficient batch-stream integrated retrieval and computation systems to optimize resource utilization.

Challenges Addressed

• Enhancing personalized ranking by fully utilizing multimodal data and increasing model capacity.
• Developing next-generation retrieval algorithms for massive-scale candidate libraries.
• Improving semantic understanding of complex and polysemous queries.
• Optimizing resource usage within search system architectures to balance performance and efficiency.
• Implementing large model technologies to significantly refine semantic understanding, retrieval precision, and resource utilization.

Qualifications

• Doctorate degree in Artificial Intelligence, Computer Science, Natural Language Processing, Computer Vision, or related fields preferred.
• Publication record in leading AI conferences or extensive research experience is advantageous.
• Strong foundation in machine learning and deep learning algorithms, with proficiency in programming languages such as C/C++ or Python.
• High intellectual curiosity, confidence, and a passionate commitment to technological innovation.
• Excellent communication and teamwork abilities.

About ByteDance and TikTok

ByteDance's TikTok is a global leader in short-form mobile video, aiming to inspire creativity and bring joy worldwide. The company values curiosity, humility, resilience, and continuous learning, fostering an inclusive environment that celebrates diverse perspectives. Offices are located worldwide including Singapore, Los Angeles, New York, London, and more.

Diversity & Inclusion

ByteDance is dedicated to cultivating an inclusive workplace where varied skills and experiences are valued. The company promotes diversity and strives to reflect the global communities its platforms serve.