WhatsApp Group Chat Analysis (Influencer and dependents)

[hakandurgut]

Hi, I have exported a WhatsApp Group Chat, there are 7 participants in this WhatsApp group.

Over a span of 218 days, a total of 7 individuals engaged in a lively conversation, resulting in a count of 3,719 chat messages exchanged.

To begin, I would like to identify the individual conversations that took place, considering that there are several instances where conversations were separated by significant periods of waiting time. These conversations can be recognized as distinct exchanges that started and ended between specific individuals.

During the span of 218 days, I noticed a pattern where certain individuals consistently responded to messages from specific people. In other words, there seemed to be a reliance on particular individuals to initiate conversations before others joined in. I must emphasize that this observation is based on my intuition and personal perception rather than scientific evidence. Nevertheless, I am eager to find a way to substantiate and provide evidence for this observation.

the participant names are
af
ct
eo
hq
ma
oa
sk

az chat export analysis.knwf (220.2 KB)

Could someone please assist me in solving this problem? I would greatly appreciate any help or guidance that can be provided.

I would like to express my gratitude in advance to anyone who is willing to offer their assistance. Your support is greatly appreciated.

[AlexanderFillbrunn]

Hi,
I see you already found the Lag Column, so it is rather easy. Quick note: you do not need the Lag Column for the Date&Time column, because the Date&Time Difference node has a “previous row” option. But you can lag the “Person” column and then filter by the Date&Time Difference to the previous row so that rows where the lag goes across two conversations are removed. Then you can use GroupBy to group by Person and Person (-1) and count the occurrence.
Does that help?
Kind regards,
Alexander

[HansS]

Hi @hakandurgut

I was thinking of something like this az chat export analysis v2.knwf (546.3 KB). First I identified the different conversations (a new conversation starts when there are no messages posted in de last 10 hours: but that is up to you). Second I analysed every conversation. Finding out who started it and how many messages each person contributed to this conversation.

gr. Hans