I'm currently a PhD student and graduate research assistant at the University of Michigan. I'm working with Prof. Rada Mihalcea and the LIT lab. I'm working on a project related to question answering and conversational agents for the purpose of automated academic advising.
My focus has been on (1) natural language understanding for recognizing entities relevant to dialog management and the sentiment toward those entities and (2) AMR semantic parsing using world knowledge.
In the past I worked on information retrieval and natural language processing techniques to compile and analyze information about interests in computer science from three sources: ACM publications, statements of purpose, and survey data. I studied how interests in computer science topics vary across gender, age, level of experience, and a few other attributes.
I received my Bachelors of Science in Computer Science from Kettering University. There I worked with Prof. Saroja Kanchi on efficient localization of wireless sensor networks and with Prof. Peter Stanchev on classification of paintings using MPEG-7 descriptors.
My current research interests are related to semantic parsing and supervised learning on dialog subtasks in human-human corpora.
PapersWelch, Charlie, Kummerfeld, Jonathan K, Feng, Song, & Mihalcea, Rada (2018). World Knowledge for Semantic Parsing with Abstract Meaning Representation. In 11th Language Resources and Evaluation Conference (LREC).
DatasetsWelch, Charlie, & Mihalcea, Rada (2017). Targeted Sentiment in Student Comments. This is a dataset consisting of student comments annotated for targeted sentiment. The dataset includes annotations of courses and instructors mentioned in student comments, as well as annotations of the student's sentiment toward those entities. A total of 1,042 comments, along with their annotations, are included.
Bot Design - A website aimed at people who are just starting to understand computer programming. It introduces users to a scripting language that allows them to create a chat bot that they can talk to. This has been used at Girls Encoded, Xplore Engineering, Girl Day, CS Kickstart and other events aimed primarily at high school students.
Bee Focused - Versatile personal task management software that is simple to use. It currently includes functionality for creating hierarchies of tasks, viewing and sorting completed tasks, and adding notes, which are searchable. Searchable notes make it an effective research or personal notebook.