The Communication Data and Network Analytics lab (CDNA) was established in August 2020. It is a cross-disciplinary team composed of social scientists and information engineering researchers. The main goal of CDNA is to provide a comprehensive understanding of communication behaviors related to important social issues, using triangulation of different research methods. The topics being examined include: fake news detection and verification, consumer risk assessment, public health risk information dissemination, political polarization, etc. Combining traditional social scientific methods (e.g., survey, experiment) with computational approaches (e.g., texting mining), the core line of CDNA research is on online public opinion mining, as well as the assessment of the process, dynamics, and effects of human communication behaviors through the integration of user log and self-reported data from survey or experiment.

In addition to research, the CDNA team is also committed to providing academic and public services, including several data platforms for use by domestic scholars and the general public. The “online data platform" is created for scholars to search and retrieve online data relevant to the topic of their interest. Users can search the online data by keywords to understand how a topic is being reported by the news media and how online public opinions on that topic change over time. In addition, CDNA provides a “human annotation and matching platform" for academics who have the need for annotating a large amount of data. They will be able to use the platform to find annotators and complete the task. Finally, as a result of CDNA’s research projects on different social risks and political issues, a “verification platform for false and/or biased information" is developed. Users only need to enter the news material, and the platform will automatically provide an assessment of its veracity (e.g., whether a news story is misinformation) and nature (e.g., whether an article is advertorial or whether it contains specific political stance).

In addition to research projects and service platforms, CDNA regularly holds academic lectures, covering diverse topics related to texting mining, data analysis, AI and other latest research development. CDNA has also expanded its perspective to the world, actively establishing a transnational scholarly community.

Project

The main goal of CDNA is to provide a comprehensive understanding of communication behaviors related to important social issues, using triangulation of different research methods. The topics being examined include: fake news detection and verification, consumer risk assessment, public health risk information dissemination, political polarization, etc. Combining traditional social scientific methods (e.g., survey, experiment) with computational approaches (e.g., texting mining), the core line of CDNA research is on online public opinion mining, as well as the assessment of the process, dynamics, and effects of human communication behaviors through the integration of user log and self-reported data from survey or experiment.

Platform

Online Data Service Platform

Provides keyword search to get a sense of the volume and the

content related to the topic of interes.

Offers quick analysis (e.g., sentiment analysis) and visualization of the

topic trend).

CNDAcrawlerHUB

The results of scheduled web scraping for news and other online data will be classified by time and topics.

Researchers will be able to search using keywords, providing them with data ready for analysis, thus avoiding the various difficulties that may arise with web scraping. Each dataset will display the first 50 entries for preview.

Researchers interested in the full dataset are encouraged to request access by contacting our center via email.

(Under construction).

Online Panel

Data annotating is the first step of conducting machine-learning.

This matching platform is to provide scholars a platform to find appropriate annotators.

Public Platform for Information Verification

Provides the public a way to determine the veracity of a news story.

Serves as a go -to place for the public to detect the quality of online information related to consumer risks (advertorials), political orientation, etc.