Monica Rogati

Monica Rogati

Senior Data Scientist @ LinkedIn

Location
Sunnyvale, California (San Francisco Bay Area)
Industry
Internet

As a LinkedIn member, you'll join 135 million other professionals who are sharing connections, ideas, and opportunities. And it's free! You'll also be able to:

  • See who you and Monica Rogati know in common
  • Get introduced to Monica Rogati
  • Contact Monica Rogati directly

View Full Profile

Monica Rogati's Overview

Current
  • Senior Research Scientist - Analytics at LinkedIn
Past
  • Research Intern at AT&T Research
  • Undergraduate Teaching Assistant & Software Developer at University of New Mexico
  • Logic Design Assistant at NASA Microelectronics Research Center @UNM
Education
  • Carnegie Mellon University
  • Carnegie Mellon University
  • The University of New Mexico
  • Tudor Vianu Computer Science High School / Liceul de Informatica
Recommendations

7 people have recommended Monica

Connections

500+ connections

Websites

Monica Rogati's Summary

Hands on, no-nonsense data science generalist; pioneered data driven products with multi-million dollar business impact. Instrumental in building the LinkedIn product analytics team from 2 to 10 data scientists.

Strong background in applied machine learning (CMU CS PhD), social network analysis, recommender systems, statistical text mining and multilingual information retrieval -- and a knack for branding.

I turn data into products, actionable insights and (news) stories.

Specialties

applied machine learning, recommender systems, analytics, data science, data driven products, social network analysis, crowdsourcing, multilingual information retrieval, text classification, statistical machine translation, fraud detection, big data, large-scale applied text mining and data mining, evaluation design, medical applications, statistical natural language processing, clustering, event detection and tracking, personal finance, tomatoes.

Monica Rogati's Experience

Senior Research Scientist - Analytics

LinkedIn

Public Company; 1001-5000 employees; LNKD; Internet industry

May 2008Present (3 years 9 months)

How do you know Darth Vader’s LinkedIn account is not real? How do you train a recommender system *before* it goes live? What are the hottest companies this year, and more importantly, how do you leverage terabytes of data to define “hot”? What algorithms will find your next sharp CTO in the 10^8 LinkedIn profile haystack? I answer these types of questions every day - and more importantly, I ask them.

The hats: applied machine learning scientist, hands-on prototyper, product manager, data miner, talent high-pass filter, mentor, spokesperson, fraud fighter, crowdsourcer, acquisition due diligence, namer … and distributed database guinea pig.

Impact:

✔ Created and implemented the initial LinkedIn job-to-candidate matching system (Talent Match), now a multi-million dollar product.
✔ Developed the first machine learning model for LinkedIn’s People You May Know, immediately doubling invitation rates.
✔ Created and implemented the "Groups You May Like" product (v1.0) resulting in a double digit engagement rate.
✔ Coaxed stories out of terabytes of raw data -- they brought us to Good Morning America, the front page of the Wall Street Journal, The Economist and hundreds of other media outlets. This includes the top shared story in LinkedIn’s history (on overused buzzwords), job title trends, successful first names, “hot” companies and industries, labor market movements etc.
✔ Pioneer, scientist, and evangelist for the LinkedIn’s recommender system line of products: Jobs You May Be Interested In, Similar Jobs, Talent Match etc. Technical contributions include: solving the cold start training problems, a generalized metrics / evaluation framework, feature engineering, model building & visualization.
✔ Crowdsourcing specialist (Mechanical Turk, Crowdflower, Samasource).
✔ Fraud detection (fake accounts, spam, TOS violations).
✔ Ran the analytics summer intern program (sourcing, interviewing, on-boarding and mentoring).

Educational Institution; 5001-10,000 employees; Higher Education industry

September 2000May 2008 (7 years 9 months)

Advisors: Yiming Yang and Jaime Carbonell. Examined feature selection for text classification and cross-lingual information retrieval using parallel corpora, including full system implementation. Participated in CLEF 2001 and 2003. Supervised two undergraduates working on web mining for CLIR resources. ♦ Supervised an MS student working on Named Entities for Topic-Conditioned Event Tracking. ♦ Thesis work on domain adaptation of translation models, as applied to Cross-Language Information Retrieval, Topic Conditioned Event Tracking and Machine Translation. ♦ Crucial role in the CMU GALE Distillation group: proposal writing with advisors, end-to-end system design and specification, multi-site coordinator (CMU, UPitt, IBM), user study coordination and design with UPitt, evaluation dataset design.

Public Company; 10,001+ employees; IBM; Information Technology and Services industry

May 2005August 2005 (4 months)

Public Company; 10,001+ employees; IBM; Information Technology and Services industry

May 2003August 2003 (4 months)

Supervisor: Scott McCarley. Unsupervised learning for Arabic stemming. Work published: ACL'03.

Research Intern

AT&T Research

Public Company; 10,001+ employees; T; Telecommunications industry

May 2000August 2000 (4 months)

Supervisor: Marilyn Walker. Worked towards improving the DARPA Communicator by treating sentence planning as a stochastic learning problem. Work published : NAACL'01, ACL'01.

Undergraduate Teaching Assistant & Software Developer

University of New Mexico

Educational Institution; 10,001+ employees; Higher Education industry

August 1997December 1999 (2 years 5 months)

Teaching Assistant, UNM CS Department (1998)
Supervisor: Barak Pearlmutter. Responsible for holding a weekly recitation section for a Non-Imperative programming (Scheme) class. Duties included grading, lecturing, holding office hours and assisting with exam questions. Developed an automatic grader program to evaluate student projects and e-mail students a progress report.
Software Developer, UNM CS Department (1998)
Leader of a two-person team in charge of developing a web-based grade database (SQL, C++) to be used for several courses in the Computer Science Department, later used by the three largest classes in the department.
Recitation Instructor, UNM CS Department (1997)
Held a weekly recitation section for a beginning C++ based Computer Science class.

Logic Design Assistant

NASA Microelectronics Research Center @UNM

19961999 (3 years)

Developed small logic design projects, including a packetizing chip software simulation and a multiplexer usage optimization for multiple outputs boolean functions; disassembled and redesigned a microcontroller test bench.
Responsible for student assignments evaluation for a logic design class.

Monica Rogati's Education

Carnegie Mellon University

Ph.D., Computer Science

20032008

Thesis:
Domain Adaptation of Translation Models for Multilingual Applications

Supervisors: Yiming Yang and Jaime Carbonell.
Examined feature selection for text classification and cross-lingual information retrieval using parallel corpora, including full system implementation. Participated in CLEF 2001 and 2003. Supervised two undergraduates working on web mining for CLIR resources. ♦ Supervised an MS student working on Named Entities for Topic-Conditioned Event Tracking. ♦ Thesis work on domain adaptation of translation models, as applied to Cross-Language Information Retrieval, Topic Conditioned Event Tracking and Machine Translation. ♦ Crucial role in the CMU GALE Distillation group: proposal writing with advisors, end-to-end system design and specification, multi-site coordinator (CMU, UPitt, IBM), user study coordination and design with UPitt, evaluation dataset design.

Carnegie Mellon University

MS, Computer Science

20002003

TA'd Danny Sleator's Data Structures and Algorithms (Java). Designed exam questions and assignments, lectured to a class of 150+.

The University of New Mexico

BS, Computer Science

19962000

First in my graduating class (summa cum laude)
Undergraduate research in neural networks and NLP.
Undergraduate TA for functional programming (Scheme).

Tudor Vianu Computer Science High School / Liceul de Informatica

Computer Science

Monica Rogati's Volunteer Experience & Causes

  • Volunteer Interests

    • Causes I care about:

      • Economic Empowerment
      • Education
      • Science and Technology
    • Organizations I support:

      • Kiva.org
      • Oxfam
      • DonorsChoose.org

Monica Rogati's Patents

  • Trainable Sentence Planning System

    • United States Patent 7,729,918
    • Issued June 1, 2010
    Inventors: Monica Rogati, Owen Christopher Rambow (2nd), Marilyn A. Walker (1st)

Monica Rogati's Publications

Monica Rogati's Additional Information

Websites:
Interests:

Large-scale applied text mining and data mining, including domain-specific applications, multi-lingual information access and organization, evaluation design, statistical methods for text processing, text classification, finance and medical applications, applied machine learning, statistical natural language processing, information retrieval, event detection and tracking, machine translation.

Groups and Associations:
Honors and Awards:

Outstanding Junior & Senior of the Year
CRA Outstanding Undergraduate 2000 Honorable Mention
UNM university-wide commencement speaker
Regional ACM programming contest: 1st & 2nd place (undergrad)
Full ride+ scholarship from the NASA Microelectronics Research Center

Contact Monica for:

  • career opportunities
  • consulting offers
  • new ventures
  • expertise requests
  • getting back in touch

View Monica Rogati’s full profile to...

  • See who you and Monica Rogati know in common
  • Get introduced to Monica Rogati
  • Contact Monica Rogati directly

View Full Profile

Viewers of this profile also viewed...