Algorithm Engineer

Duties include:

1. Research and develop semantic understanding algorithms for search engine scenarios, including but not limited to: user query intent recognition and semantic parsing, search term and file/entity semantic relevance modeling, semantic generalization and correction for long-tail search scenarios, and semantic understanding technology for language/multimodal search.

2. Build large-scale pre-training models, graph neural networks, and other cutting-edge technologies to enhance the semantic accuracy of search results.

3. Analyze user search behavior data, uncover semantic understanding bottlenecks and propose optimization solutions.

4. Collaborate with search sorting and knowledge graph teams to push algorithms to the online system.

5. Follow up on the latest developments in the NLP field (such as Prompt Learning, contrastive learning, etc.), and explore their application in search scenarios.

Position requirements

1. Computer-related professional master's degree or above, with at least 3 years of NLP algorithm experience.

2. Proficient in Python and deep learning frameworks (PyTorch/TensorFlow).

3. Have search engine related experience, familiar with Query Understanding/Intent Recognition/Semantic Matching, etc. scenes.

4. Familiar with BERT, Transformer, etc. models, with massive data training and optimization experience.

5. Have the following experience is preferred:

- Actual participation in the development of a search semantic understanding system

- Familiar with Elasticsearch / Lucene and other search technologies

- Have experience in large model fine-tuning or distributed training