%0 Journal Article %J Journal of Internet Services and Applications %D 2017 %T Core-periphery communication and the success of free/libre open source software projects %A Kevin Crowston %A Shamshurin, Ivan %K Apache Software Foundation %K communication %K core and periphery %K free/libre open source software (FLOSS) %K inclusive pronouns %K natural language processing %K project success %X We examine the relationship between communications by core and peripheral members and Free/Libre Open Source Software project success. The study uses data from 74 projects in the Apache Software Foundation Incubator. We conceptualize project success in terms of success building a community, as assessed by graduation from the Incubator. We compare successful and unsuccessful projects on volume of communication and on use of inclusive pronouns as an indication of efforts to create intimacy among team members. An innovation of the paper is that use of inclusive pronouns is measured using natural language processing techniques. We also compare the volume and content of communication produced by core (committer) and peripheral members and by those peripheral members who are later elected to be core members. We find that volume of communication is related to project success but use of inclusive pronouns does not distinguish successful projects. Core members exhibit more contribution and use of inclusive pronouns than peripheral members. %B Journal of Internet Services and Applications %V 8 %G eng %U http://rdcu.be/uguP %N 10 %R 10.1186/s13174-017-0061-4 %> https://floss.syr.edu/sites/crowston.syr.edu/files/170707%20JISA%20final.pdf %0 Conference Proceedings %B American Society for Information Science and Technology (ASIST) Annual Conference %D 2010 %T Machine Learning and Rule-Based Automated Coding of Qualitative Data %A Kevin Crowston %A Xiaozhong Liu %A Allen, Eileen E. %A Heckman, Robert %K FLOSS %K NLP %X Researchers often employ qualitative research approaches but large volumes of textual data pose considerable challenges to manual coding. In this research, we explore how to implement fully or semi-automatic coding on textual data (specifically, electronic messages) by leveraging Natural Language Processing (NLP). In particular, we compare the performance of human-developed NLP rules to those inferred by machine learning algorithms. The experimental results suggest that NLP with machine learning can be an effective way to assist researchers in coding qualitative data. %B American Society for Information Science and Technology (ASIST) Annual Conference %C Pittsburgh, PA %8 10/2010 %> https://floss.syr.edu/sites/crowston.syr.edu/files/ml_nlp.pdf %> https://floss.syr.edu/sites/crowston.syr.edu/files/ASIST%20poster%202p%20final.pdf