Richard Wang Email & Phone Number

Minneapolis, MN, US

• Led a dynamic team of ~20 ML engineers and data scientists, specializing in enhancing the e-commerce search engine with state-of-the-art data science and machine learning techniques.
• Architected and implemented cutting-edge solutions for critical search engine components, including query intent understanding, query expansion, synonym detection, product relevance ranking, and product desirability.
• Led the continuous enhancement of search capabilities, delivering exceptional user experiences and driving significant business outcomes, resulting in an annual incremental revenue increase of over $400 million.

Minneapolis, MN, US

• Developed and deployed advanced AI and machine learning solutions to address natural language processing (NLP) challenges at Target, leveraging cutting-edge information extraction and retrieval techniques.
• Engineered a CNN n-gram-based text classification model that substantially improved e-commerce search precision and recall, consistently achieving rates above 90%.
• Drove significant improvements in search accuracy by refining the understanding of merchandise categories and attributes through sophisticated query understanding techniques.
• Partnered with the customer support team to automate the detection and analysis of key customer feedback elements, including sentiment analysis, issue identification, and store mentions, utilizing advanced NLP.

Sunnyvale, CA, US

• Led a knowledge search team of 10+ members, focusing on medical search queries.
• Built a state-of-the-art medical knowledge graph containing over 200 million facts, enhancing system understanding of medical queries to 90% with 95% accuracy.
• Increased Click-Through Rate (CTR) of medical-related ads by 0.5% and Revenue Per Mille (RPM) by 1.5%, resulting in an annual revenue increase of $15M USD.
• Developed a virtual medical doctor chatbot capable of understanding users' medical conditions through intent detection and entity recognition.
• Enabled the chatbot to interact with users across multiple conversations using dialogue state tracking and provide medical advice through Bayesian inference and random walk techniques on the medical knowledge graph.
• Contributed to various NLP tasks, including the development of a conversation dialogue framework for an Alexa-like project, improving language model accuracy, and implementing automatic capitalization for speech.

• Founded a company focused on deep learning and understanding of online articles by automatically identifying people, places, and things mentioned.
• Provided relevant information to publishers, advertisers, and users through proprietary NLP, AI, and knowledge graph technology.
• Successfully launched on numerous publishers' websites in 2013, reaching over 1M+ daily users.
• Achieved a mouse-hover rate of 20%, indicating 1 out of 5 users engaged with our annotations.
• Accomplished an astonishing re-hover rate of 80%, showing high user engagement with 4 out of 5 users who hovered once doing it again.

Mountain View, CA, US

• Lead a project aimed at expanding the coverage and completeness of Google's Knowledge Graph, which contains over 44 million topics and 2.4 billion facts with 99% accuracy.
• Utilized unsupervised clustering and semi-supervised learning on semi-structured web data to increase Knowledge Graph's coverage and completeness by more than 5%.
• Worked on the Google Squared project, an automatic instance extraction system for categorical user queries (e.g., car makers, US presidents).
• Implemented and integrated the SEAL system from my thesis into Google Squared, improving its precision and recall by more than 30%.

Pittsburgh, PA, US

• Conducted research on machine learning, natural language processing, and information extraction.
• Implemented systems for the RADAR personal assistant project and the NELL (Never-Ending Language Learning) project at CMU.

Pittsburgh, PA, US

• Teaching assistant for the course "Advanced Natural Language Processing Seminar" at CMU.

Pittsburgh, PA, US

• Graded student's assignments and exams, made exam questions, and held office hours for the course "Information Retrieval" at CMU.

Pittsburgh, PA, US

• Developed a system for the RADD machine translation project at CMU that crawls the web and automatically identifies and extracts parallel corpora from online web pages for training machine translation models.

Mountain View, CA, US

• Developed a generic learning and extraction system that automatically produces complex and well-formulated regular expressions for extracting information learned from annotated training data.

Pittsburgh, PA, US

• Developed a system that automatically extracts cell/protein names from image captions in biological journal articles, to use as features for training a cell/protein image classifier.

Ph.D., Computer Science, Language Technologies

M.S., Computer Science, Language Technologies

B.S., Computer Science