u*********1
发帖数: 11 | 1 Responsibilities:
Participate in the AI research and application development of JD.com,
including (but not limited to) intelligent robot, NLP, risk control, etc.
responsible for team-leading on research of new topics and critical issues
solving. Guide the technical development and member-training of the team,
management skills required.
Basic Qualifications:
• MS/PhD Degree of CS, graduation from outstanding colleges is
preferred, (Stanford, Carnegie Mellon, UC Berkeley, etc.)
• Minimum 5+ years of R&D experience. Candidates must possess
technologies relevant to big data processing, i.e., machine learning, neural
network algorithms etc.; must show deep research in at least one of the
following topics: NLP, Data Mining, Big Data Reasoning, Risk Control or
other big data applications.
• Broadview on the latest development of technology worldwide,
especially in artificial intelligence (research/papers/technical update);
capability of transmitting the technology into the production stage.
• Insights of technical development tendencies in related research
domains.
• Familiar with distributed systems, statistics, optimization, with
excellent programming background and engineering experience.
• High executive ability to overcome all possible challenges
towards the target.
• AI work experience in Google/Facebook/etc. is a plus.
Research topics:
If you join us, we have the following topics:
Topic 1: Semantic Query Intention Recognition
Analyze the semantic representations of character/word/sentence, make
accurate analysis between user’s question and existing question/answer
pairs set, to accurately split user’s questions and answers to proper small
parts, recognize user’s accurate intention through analysis of semantic
structure.
Through the research on semantic representation of all characters/words/
sentences, analysis of current existing pairs of users’ questions and
answers, segmentation of these Q&A into elements and semantic structure
analysis, implement accurate intention recognition on user’s query,
Topic2: Contextual Query Intention Analysis
Analyze and integrate several rounds of talk, analyze beginning and ending
sentence for each of intentions from user, recognize sentence segmentation,
exclude useless sentences, improve accurate evaluation ability for
recognizing meaning for single or multiple rounds talk, and to improve
intention recognition/entity recognition accuracy rate.
Implement an integral analysis of multiple rounds of talk, analyze the
starting and ending information from each intention of the user, recognize
the sentence segmentation and filter all meaningless sentences, in order to
improve the accurate evaluation performance on meanings of the single/
multiple rounds of talk, and finally improve the intention recognition/
entity recognition accuracy rate.
Topic 3: QA Data Mining
Mine accurate question and answer pairs from massive talk corpus between
user and customer service staff. Ensure accuracy rate between question and
answer. And while user chats with a robot, find the best match from question
/answer pairs mined before, providing a good choice for robot’s answer.
Accurate mining of meaningful Q&A pairs from between users and Customer
Representative; ensure the accurate matching between questions and answers,
and assist the robot’s answer through the recognition and matching with the
current database.
Topic 4: Accurate Prediction for User’s Consultation
Through analysis of big data about user’s behavior/logs/historic
information, evaluate user’s main consultation intention before user write
down consultation questions, so to divide consultation service level
according to intention.
About JD.com, Inc
JD.com, Inc. is one of the Top 10 Internet Company, the largest online
retailer in China and the country’s largest Internet company by revenue.
The Company strives to offer consumers the best online shopping experience.
Through its content-rich and user-friendly website jd.com and mobile
applications, JD.com offers a wide selection of authentic products at
competitive prices and delivers products in a speedy and reliable manner. JD
.com has the largest fulfillment infrastructure of any e-commerce company in
China. It operates 7 fulfillment centers, 196 warehouses in 46 cities, and
4,760 delivery and pickup stations in 2,266 counties and districts across
China. JD.com is a member of the NASDAQ100 and a Fortune 500 company.
If interested, feel free to contact us: [email protected]! |
|
