LinkedIn stats
The repository contains messages that I received on LinkedIn, and some Python scripts extracting interesting stats from it.
The stats so far:
Autogenerated : 100/153 (65%) ███░░
Has my name in the message : 146/153 (95%) █████
Read my CV : 000/153 (00%) ░░░░░
Has salary range : 023/153 (15%) █░░░░
Has company name : 063/153 (41%) ██░░░
Has emojis : 036/153 (24%) █░░░░
Has email : 038/153 (25%) █░░░░
Has a link : 062/153 (41%) ██░░░
Has word "Python" : 108/153 (71%) ████░
Send on weekend : 013/153 (08%) ░░░░░
Send during working hours : 127/153 (83%) ████░
Duplicate messages : 019/153 (12%) █░░░░
Duplicate companies : 024/153 (16%) █░░░░
Average message size : 1178 chars
To reproduce:
python3 parser.py | python3 stats.pyExplanation for fields:
- "Autogenerated". My name on linkedin has an apparent "junk symbols". Namely, a SQL-injection:
' OR 1 -- .. If i receive a message with them, I consider it autogenerated. - "Read my CV". My name on LinkedIn is Nikita. My name in all other places, including CV (and it is the preferred name) is Gram. If they address me as Nikita, they haven't opened my CV.
- "Has company name". I don't have a reliable algorithm to detect it from the text, so I manually extracted all company names from all messages. So, mistakes are possible.
- "Has word 'Python'". I'm a Python engineer, and so I expect all vacancies I receive to be relevant to it. Often, I receive vacancies for PHP, Java, or whatever else.
- "Send during working hours". Anywhere between 8:00 and 18:00.
- "Duplicate messages". Number of messages from people who contacted me before. It might be a different position but the same sender.
- "Duplicate companies". How many times I was contacted about a position in the company I was contacted about before.