Zhiyuan Xu

美国 加利福尼亚 旧金山
463 位关注者 434 位好友

加入领英，查看档案

动态

立即加入，查看全部动态

工作经历和教育背景

  • Pinterest

查看Zhiyuan的完整工作经历

查看他们的职位头衔、任职时间等。

或

点击“继续加入或登录”，即表示您同意遵守领英的《用户协议》、《隐私政策》及《Cookie 政策》。

课程

  • Algorithms and Economics of the Internet

    CSCI-GA 3033-003

  • Cloud Computing

    CSCI-GA 3033-009

  • Computer Networks

    -

  • Data Structure

    -

  • Database Concept and Applications

    -

  • Design and Analysis of Algorithms

    -

  • Distributed System

    CSCI-GA 3033-002

  • Fundamental Algorithm

    CSCI-GA 1170-001

  • Natural Language Processing

    CSCI-GA2590-001

  • Open Source Tools

    CSCI-GA 2246-001

  • Operating System

    CSCI-GA 2250-001

  • Principle of Compiler

    -

  • Programming Language

    CSCI-GA 2110-001

  • Web Search Engines

    CSCI-GA 2580-001

所做项目

  • Name Entity Recommendation System

    - Extracted name entities with Stanford NER and indexed the co-occurrence relation between NEs
    - Implemented in-document reference resolution based on textual feature and improved cross-document reference resolution results with Freebase API
    - Built inverted index for terms in NEs to provide search recommendations for ambiguous queries

    其他创始人
    查看项目

  • Sport Event Result Extraction for NBA Games

    - Developed hand-coded word lists and chunk patterns based on POS tags for sport events
    - Improved the algorithm when chunk patterns cannot fit precisely, and normalized the result
    - Achieved 0.9 in recall rate and 0.95 in precision rate on source text from ESPN, Yahoo sports and NBA team pages on facebook.

  • Pipeline for Multi-lingual Business Intelligence Service

    -

    - Built a client page using PHP, and implemented the communication between client and server using Thrift
    - Implemented the handling and processing logic of the server using C++, did the processing and analysis of text data using shell script and Python
    - Improved analysis result by applying a rank aggregation algorithm. Reduced 80% overall processing and analysis time by optimizing Hive query and utilizing multi-thread. Used shell script to reduce memory need during text processing in…

    - Built a client page using PHP, and implemented the communication between client and server using Thrift
    - Implemented the handling and processing logic of the server using C++, did the processing and analysis of text data using shell script and Python
    - Improved analysis result by applying a rank aggregation algorithm. Reduced 80% overall processing and analysis time by optimizing Hive query and utilizing multi-thread. Used shell script to reduce memory need during text processing in word segmentation and dependency parsing
    - Skills&Tools: C++, Thrift, Hive, Shell script, Python, Chasen

  • Microblog Recommendation System Based on Topic Modeling

    -

    - Colleted source text data by using Sina Weibo API and processed the source text data
    - Built topic vector for every microblog by using topic modeling tool Mallet
    - Calculated similarity of topic vectors and combined results with personalized factors in a linear model
    - Skills&Tools: Java, SQL, Mallet Topic Modeling Tool.

  • Videos Upload and Rating System Based on Amazon AWS Platform

    -

    - Developed a web application for uploading and rating video files
    - Saved videos uploaded by users in AWS S3 and rating information in AWS RDS
    - Deployed application on AWS Elastic Beanstalk and offer streaming by using AWS
    　　 CloudFront
    - Skills&Tools: Java, JSP, Amazon AWS.

  • WeNote --- Note Sharing and Searching Plugin for Chrome

    -

    - Implemented APIs processing database read and write requests from the plugins on the AWS platform
    - Collaborated with teammates to combine front-end code and back-end code
    - Skills&Tools: Java, Amazon AWS, Javascript.

    其他创始人

语言能力

  • Chinese

    母语或精通双语

  • English

    中高级 (商务会话)

Zhiyuan的更多动态

查看Zhiyuan的完整档案

  • 浏览共同好友
  • 请求引荐
  • 直接联系Zhiyuan
加入领英，查看完整档案

其他相似会员

学习在线课程，新技能轻松 get！

查看全部课程