Data Scientist, Research

Responsibilities

TikTok is the leading destination for short-form mobile video. We have several global offices including Los Angeles, New York, Austin, London, Paris, Berlin, Dubai, Singapore, Jakarta, Seoul, and Tokyo.​

At TikTok, our people are humble, intelligent, compassionate and creative. We create to inspire - for you, for us, and for more than 1 billion users on our platform. We lead with curiosity and aim for the highest, never shying away from taking calculated risks and embracing ambiguity as it comes. Here, the opportunities are limitless for those who dare to pursue bold ideas that exist just beyond the boundary of possibility. Join us and make impact happen with a career at TikTok.

About the team

The success of TikTok's data business model hinges on the supply of a large volume of high quality labeled data that will grow exponentially as our business scales up. However, the current cost of data labeling is excessively high. The Data Solutions team is built to understand data strategically at scale for all Global Business Solution (GBS) business needs. Data Solutions Team uses quantitative and qualitative data to guide and uncover insights, turning our findings into real products to power exponential growth. Data Solutions Team responsibility includes infrastructure construction, recognition capabilities management, global labeling delivery management.

Responsibilities

- Lead technical research on bridging the understanding between human and machine learning.

- Build the library of elementary concepts that serve as factors to form content

- Interpretation of natural signals (image, audio, text, video) into structured computational signals for ML models to process like the human brain

- Identify patterns that indicate intentions, and verify them with experiments

- Incorporate a variety of statistical and machine learning techniques - such as logistic regression, clustering, mixed modeling, decision trees and neural networks - on multi-modal datasets

- Understand underlying data sources and their limitations. Create innovative approaches to answer pressing questions, prepare complex data analyses and models that help solve issues, drive the scaling of automated processes and deliver significant measurable impact

- Communicate with machine learning engineers and product partners to understand business needs and provide analytical solutions

- Act as an analytics translator, communicating complex data insights through exploratory analysis and research to discover potential bottlenecks to suggest improvement and workflow of internal teams

Qualifications

- Advanced degree in social, behavioral, human-centric interaction or cognitive sciences (Linguistics, Sociology, Anthropology, Psychology) or other research-oriented disciplines within Humanities. Experience in quantitative fields (Computer Science, Engineering, Statistics, Mathematics or related fields) is a plus, Ph.D. is a plus.

-Experience in Graph Neural Network(GNN) or NLP will be highly advantageous.

- Proven experience working with large datasets and relational databases (Hive, SQL)

- 3-7 years of hands-on behavioral/ cognitive research experience in a business environment

- Experience in statistics and experimental design and data mining techniques (k-means clustering, regression, decision trees, clustering, neural networks, etc.)

- Experience in programming computational and statistical algorithms for large data sets.

- Proficiency in Python packages such as pandas, seaborn, scikit-learn, dplyr or nltk

- Distinctive communications skills and ability to communicate analytical and technical content in an easy-to-understand way to both technical and non-technical audiences.

- Intellectual curiosity, along with excellent problem-solving and quantitative skills, including the ability to disaggregate issues, identify root causes and recommend solutions