RP-commenting-practices-multiple-sources

Replication package for the paper "What do Developers Discuss about Code Comments?"

Structure

Appendix.pdf
Tags-topics.md
Stack-exchange-query.md

RQ1/
  LDA_input/
    combined-so-quora-mallet-metadata.csv
    topic-input.mallet

  LDA_output/
    Mallet/
      output_csv/
        docs-in-topics.csv
        topic-words.csv
        topics-in-docs.csv
        topics-metadata.csv
      output_html/
        all_topics.html
        Docs/
        Topics/

RQ2/
  datasource_rawdata/
    quora.csv
    stackoverflow.csv
  manual_analysis_output/
    stackoverflow_quora_taxonomy.xlsx

Contents of the Replication Package

• Appendix.pdf- Appendix of the paper containing supplement tables

• Tags-topics.md tags selected from Stack overflow and topics selected from Quora for the study (RQ1 & RQ2)

• Stack-exchange-query.md the query interface used to extract the posts from stack exchnage explorer.

• RQ1/ - contains the data used to answer RQ1

  • LDA_input/ - input data used for LDA analysis
  • combined-so-quora-mallet-metadata.csv - Stack overflow and Quora questions used to perform LDA analysis
  • topic-input.mallet - input file to the mallet tool
  • LDA_output/
  • Mallet/ - contains the LDA output generated by MALLET tool
    • output_csv/
      • docs-in-topics.csv - documents per topic
      • topic-words.csv - most relevant topic words
      • topics-in-docs.csv - topic probability per document
      • topics-metadata.csv - metadata per document and topic probability
    • output_html/ - Browsable results of mallet output
      • all_topics.html
      • Docs/
      • Topics/

• RQ2/ - contains the data used to answer RQ2

  • datasource_rawdata/ - contains the raw data for each source
  • quora.csv - contains the processed dataset (like removing html tags). To know more about the preprocessing steps, please refer to the reproducibility section in the paper. The data is preprocessed using Makar tool.
  • stackoverflow.csv - contains the processed stackoverflow dataset. To know more about the preprocessing steps, please refer to the reproducibility section in the paper. The data is preprocessed using Makar tool.
  • manual_analysis_output/
  • stackoverflow_quora_taxonomy.xlsx - contains the classified dataset of stackoverflow and quora and description of taxonomy.
    • Taxonomy - contains the description of the first dimension and second dimension categories. Second dimension categories are further divided into levels, separated by | symbol.
    • stackoverflow-posts - the questions are labelled relevant or irrelevant and categorized into the first dimension and second dimension categories.

      - quota-posts - the questions are labelled relevant or irrelevant and categorized into the first dimension and second dimension categories.

Frequently used terms and phrases in various Program Guidelines and Applications. For additional information, visit the department Funding page: https://www.austintexas.gov/department/economic-development/funding

This dataset contains 70,427 cross-linked Twitter-[GHTorrent](http://ghtorrent.org) user pairs identified as likely belonging to the same users. The dataset accompanies our research paper:

@inproceedings{fang2020tweet,

author = {Fang, Hongbo and Klug, Daniel and Lamba, Hemank and Herbsleb, James and Vasilescu, Bogdan},

title = {Need for Tweet: How Open Source Developers Talk About Their GitHub Work on Twitter},

booktitle = {International Conference on Mining Software Repositories (MSR)},

year = {2020},

pages = {to appear},

publisher = {ACM},

}

The data cannot be used for any purpose other than conducting research.

Due to privacy concerns, we only release the user IDs in Twitter and GHTorrent, respectively. We expect that users of this dataset will be able to collect other data using the Twitter API and GHTorrent, as needed. Please see below for an example.

To query the Twitter API for a given user_id, you can:

• Apply for Twitter developer account here.

• Create an APP with your Twitter developer account, and create “API key” and “API secret key”.

• Obtain an access token. Given the previous

curl -u "

The response looks like this: {"token_type":"bearer","access_token":"<...>"}

Copy the "access_token".

• Given the previous access token, run:

curl --request GET --url "https://api.twitter.com/1.1/users/show.json?user_id=

The GHTorrent user ids map to the users table in the MySQL version of GHTorrent. To use GHTorrent, please follow instructions on the GHTorrent website.

This is the dataset used in the respective research work. The abstract is available below.

If you want to cite this work, please use:

Georgia M. Kapitsaki, Maria Papoutsoglou, Daniel German and Lefteris Angelis, What do developers talk about open source software licensing?, to appear in the Proceedings of the Euromicro Conference on Software Engineering and Advanced Applications, SEAA 2020.

Free and open source software has gained a lot of momentum in the industry and the research community. Open source licenses determine the rules, under which the open source software can be further used and distributed. Previous works have examined the usage of open source licenses in the framework of specific projects or online social coding platforms, examining developers specific licensing views for specific software. However, the questions practitioners ask about licenses and licensing as captured in Question and Answer websites also constitute an important aspect toward understanding practitioners general licenses and licensing concerns. In this paper, we investigate open source license discussions using data from the Software Engineering, Open Source and Law Stack Exchange sites that contain relevant data. We describe the process used for the data collection and analysis, and discuss the main results. Our results indicate that clarifications about specific licenses and specific license terms are required. The results can be useful for developers, educators and license authors.

Economic Development Department Definition Guide