Datasets and examples¶

Ideas for real-world applications of truth discovery, to use as examples or to try and actually find data for.

• From [LLG+14]:

  • Customer information can be found from multiple databases in a company

  • A patient’s medical records may be scattered across different hospitals

  • Construction of indoor floorplans via social sensing – e.g., accelerometer, gyroscope, and compass sensor measurements from smartphones as users walk around. Goal is to measure distances between indoor points; conflicts arise due to different walking patterns etc. See references [1], [28] in [LLG+14].

• Some social sensing examples from [WKLA12]:

A classical example is geotagging campaigns, where participants report locations of conditions in their environment that need attention (e.g., litter in public parks).

Examples include documenting the quality of roads [25], the level of pollution in a city [20], or reporting garbage cans on campus [24].

• Population and biographies datasets from [PR10]

• Fake News Challenge data might have useful data. It seems like there are some datasets available in their GitHub repos.

• Looks like there a few datasets in [ZhangShengLiWu18] which are available online (I have not read the paper)