We are trying to extract an event log from chats between users in a Jabber server (OpenFire). In the extracted event log we just have two activities (chat, replayto). Our purpose is to generate a social network from the event logs using ProM tool. This social network will be used to calculate an influence metric.
If the event log is intended for social mining, is it important the number of activities? Should I extract the social network directly from the OpenFire logs?
That is a good question. I think that if we only have two activities with the objective of inferring a social network between the actors it would be correct to extract an event log. The most important is to know the intensity of the relationship, that is, the number of times the activities are executed, which would be the process instances. This could be a key element in determining influence among network actors.
That is a good question. I think that if we only have two activities with the objective of inferring a social network between the actors it would be correct to extract an event log. The most important is to know the intensity of the relationship, that is, the number of times the activities are executed, which would be the process instances. This could be a key element in determining influence among network actors.
it is the whole of the activities that is important. In any case if it is reduced to two elements it is not a problem . These two activities can be used to generate relationships between the different actors.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts