CCI Faculty Members An, Chatterjee and Pirmann Among 2021-2022 Drexel Provost and Research Award Winners

Photos of faculty members Pirmann, Chatterjee and An
From left to right: Tammy Pirmann, EdD; Preetha Chatterjee, PhD; and Yuan An, PhD

Three College of Computing & Informatics (CCI) faculty members were among the recipients of the 2021-2022 Research, Scholarship, and Creativity Awards and the Provost Awards for Teaching, Scholarship, and Professional Service.

Presented by Drexel University’s Office of Research & Innovation and the Office of Faculty Advancement and Undergraduate Affairs, these awards support the ongoing development of Drexel’s research, scholarship, and creative activities and to recognize excellence in teaching and professional service.

Assistant Teaching Professor Tammy Pirmann, EdD received the Evidence-Based Teaching Award in Undergraduate Science, Technology, Engineering, and Math (STEM) award, which is granted annually to a full-time tenure track or teaching faculty member in a STEM department who has taught at Drexel for at least two years. Winners must show application of pedagogical practices that have been shown in the research to be beneficial for teaching students, or that they produce unique ways that are being evaluated in their own classrooms. Pirmann is a steadfast advocate for computer science (CS) education, most recently serving as co-PI on a funded grant to develop a researcher/practitioner partnership with the School District of Philadelphia to enhance their efforts to provide equitable CS education at the secondary level in all school buildings (CS4All).

Associate Professor Yuan An, PhD and Assistant Professor Preetha Chatterjee, PhD were both recipients of the Drexel Office of Faculty Affairs’ 2022 Faculty Summer Research awards, which provide tenured and tenure-track faculty the opportunity to pursue research activities that will enhance their careers as well as increase their contributions to Drexel. The awards will support summer research for both faculty members; An’s project titled “Embedding-based Alignment for the Materials Science Ontologies in the MatPortal Repository” and Chatterjee’s project titled “Towards Enabling Affective Awareness in Free Open Source Software (FOSS) Projects.” Full project abstracts are listed below.


Abstract for “Embedding-based Alignment for the Materials Science Ontologies in the MatPortal Repository” by Yuan An, PhD (PI):

Data-driven materials science has become a new paradigm of materials study. However, data heterogeneity and machine processibility remain the key bottlenecks hindering the advancement of data-driven materials discovery. It is imperative to develop semantic approaches for unifying the distributed and disparate big data to empower domain scientists.  To this end, we have embarked on a project that applies knowledge graph techniques for semantic data discovery, extraction, and integration in materials science. The research consists of several key components including ontology mapping, materials data annotation, and information extraction from unstructured scholarly articles. In this summer research project, we will focus on investigating the current materials science related ontologies in the MapPortal repository ( and developing an automatic tool to discover semantic alignments between the materials science ontologies by leveraging our OTMaponto system. 


Abstract for “Towards Enabling Affective Awareness in Free Open Source Software (FOSS) Projects” by Preetha Chatterjee, PhD (PI):

Free Open Source Software (FOSS) projects are a societal good. FOSS projects have increased the speed of digital advancement and therefore technology that relies on FOSS is pervasive (e.g., FOSS underpins nuclear submarines, industrial marketplaces, and mobile phones). FOSS projects rely on the work of volunteers who contribute their time in order to improve their own software development skills and to be part of a community. To ensure FOSS are successful, we must protect and encourage the participation of these volunteers. However, contributors often abandon projects because of social and emotional factors [1], e.g., because they experience negative emotions about the direction of the project, have a negative experience with other participants, or feel isolated and ignored by other participants. Project managers are often unaware of such occurrences due to the distributed nature of FOSS project communication (often involving many channels like issue trackers, chat, pull request comments) and the large number of contributors.

Emotions impact cognitive and communication skills, thus influencing activities which are collaborative in nature and require creativity and problem-solving skills, such as software development. Emotions (e.g., joy, anger, fear) are pervasive in daily software engineering operations, and are known to be significant indicators of work productivity and team attrition, both of which contribute to long-term project sustainability [2], [3]. For instance, positive emotions correlate with increased developer productivity, while negative emotions can lead to project attrition. Automatically measuring software developer emotions has only recently become possible with the advancements in Machine Learning (ML) and Natural Language Processing (NLP) techniques such as sentiment analysis and deep neural network models. 

Directly applying approaches from NLP to software engineering (SE) corpora is not straightforward as many have observed the distinct nature of SE text, containing words with domain-specific meanings (e.g., throws, bugs). More specifically, several studies of using general-purpose NLP corpora and techniques for extracting emotions have shown very poor results [4], [5], indicating that this problem requires bespoke approaches (e.g., training data, sentiment lexicons, pre-trained models) that are specific to software engineering and maybe even more narrowly fine-tuned to a specific software artifact (e.g., bug reports vs. pull request comments) or a software project. Measuring affect or emotion in FOSS projects is particularly challenging as the community members have diverse backgrounds, thus having different values and perspectives. Consolidating the varied perspectives becomes harder due to the presence of complex emotions and affective states of the community participants, as manifested in multiple FOSS communication channels [6].

Through the proposed project, PI Chatterjee aims to advance the state of the art by developing techniques for understanding and classifying the emotions of software developers actively participating in a FOSS software project based on their written communication. Specifically, this proposal will investigate: 1) leveraging developer utterances across different channels of a FOSS project to create models that can automatically classify participant emotion; 2) automatic tools and sources of ground truth to train and validate machine learning-based models, and conduct post hoc participant surveys. The proposed techniques will help raise the awareness of FOSS project managers so they can better respond to the expressed emotions of their project participants, and enable affective awareness among participants that could improve communication and collaboration in FOSS projects.

In This Article

Related News

Contact Us

Have a question? We’re eager to talk with you.

Contact Us