Главная
Study mode:
on
1
Intro
2
Four challenges facing Al teams at SEEK
3
About me
4
About SEEK's data
5
About AIPS
6
How my team helps hirers and candidates
7
Job data
8
Candidate data
9
Structured "bag of tags" representations
10
Unstructured vector space representations
11
Many expectations of Al
12
Delivering innovative Al services
13
Delivering FAT Al services
14
Talent Search using using Elastic ANN
15
Iterative development of a new Al service
16
Product science at SEEK
17
Many models, many pipelines?
18
Market-adapted model training
19
Cross-squad knowledge sharing
20
Cross-squad model sharing with Python
Description:
Explore the technical and non-technical challenges faced by SEEK, a global employment marketplace, in maximizing the potential of documents, natural language processing (NLP), and artificial intelligence (AI). Dive into a 42-minute keynote presentation from ADCS 2021, delivered by Terrence Szymanski. Learn about SEEK's approach to processing multinational, multilingual job ads and candidate resumes, and discover how they tackle issues such as accuracy, performance, and reliability of services. Gain insights into organizational strategies for large teams, ethical considerations in AI development, and the protection of user data. Examine SEEK's innovative solutions, including structured "bag of tags" representations, unstructured vector space representations, and the use of Elastic ANN for talent search. Understand the iterative development process for new AI services, the implementation of product science, and strategies for cross-squad knowledge and model sharing in Python.

Maximizing the Potential of Documents, NLP, and AI - Technical and Non-technical Challenges

Association for Computing Machinery (ACM)
Add to list