search

fani-lab / SEERa

A framework to predict the future user communities in a text streaming social network based on the users’ topics of interest.
Other
4 stars 5 forks source link

Users similarities problem #27

Closed soroush-ziaeinejad closed 2 years ago

soroush-ziaeinejad commented 2 years ago

The similarities between users are not calculated properly. That's why there is a strict threshold for them to be connected to each other. Thresholds lower than 0.7 make a huge community but thresholds greater than 0.75 just keep communities that their members have exactly the same connections (Which is also not a good thing! There are multiple nodes with exactly the same connections in the users' graphs).

0<Threshold<0.7: Graph3

0.75<Threshold<1 Graph3

hosseinfani commented 2 years ago

Not sure I understood. We can discuss this in our weekly meeting.