Objective: providing the echOmmunity with a clear mapping of the mapping

Data

Possible data per user:

• Country
• City
• Keywords
  • Keywords could be separated into 2 categories : roles in echOpen / skills
  • Keywords must be chosen from a pre-defined list
• Nick on slack
• Channels on slack
• People mentionned by user on slack
• Mentions of user by other people on slack
• Nick on basecamp
• Github account
• Programming skills (languages, databases, ...)

Origin of the data

• Former CSVs
• Questionnaires through a bot asking people with private messages on Slack
• Bot crawling slack to see mentions of people by others
• Bot crawling, seeing who is following what (github, slack channel)

Channels of capture: slack, github, ...?

The Slack Project

• the getInfo.py gets the content of the slack channels;
• then analyse.py creates the content of the json, per channel;
• and CreateUsers.py creates the main user_json, and creates the pages on the github pages.

Deliverables

Graphs

How ?

1. Bots crawling slack / forms --> store as .csv files
2. Python+neo4j --> update graph database
3. Python+neo4j --> queries for personnalized suggestions --> results stored as .csv files
4. Delivery :
   • Slack bot sending private message to user
   • Webpage for "global" informations about the community

What ?

• Mapping people by geography
• Mapping by interest
• Connecting

Recommendation:

Have an algo recommend you:

• to connect with the 3 most similar profiles
• to subscribe to the three most interesting channels on slack
  • Query of type : " Most popular channels among people who have your skills/interests..."
• to connect with slack members
  • Query of type : "People who have your skills/interest also interact with..."
• to follow on github the three most interesting projects for you
  • Query of type : "People who have your skills/interests contribute to..."

Doing..

Categories:

• Hardware: pcb, electronic, fpga, cpld, stm32, arduino, power, pulser, kicad,

• Software: code, android, java,

• Legal: patent, agreement, cla,

• Medical: doctor, patient,

• Design: design, user,

• Community: graph, community, communication, event, contribution, contributor, wiki

• Include reactions to link two users

Ideal json structure (one .json per channel)

{
  "channel_id" : ,
  "info": {
    "nb_users" : ,
    "community": ,
    "design": ,
    "hardware": ,
    "legal": ,
    "medical": ,
    "posts": ,
    "software":
  },
  "users": [
    id_of_user1,
    id_of_user2,
  ],
  "users_info": {
    id_of_user1: {
      "community": ,
      "design": ,
      "hardware": ,
      "legal": ,
      "medical": ,
      "posts": ,
      "software":
    },
    id_of_user2: {
      "community": ,
      "design": ,
      "hardware": ,
      "legal": ,
      "medical": ,
      "posts": ,
      "software":
    },
  },
  "mentions" : [
       {"user_id" : id_of_user1,
        "mentioned_user_id" : id_of_user2,
        "timestamp" : ts
        },
       {"user_id" : id_of_user1,
        "mentioned_user_id" : id_of_user2,
        "timestamp" : ts
        }
   ],
   "reactions" : [
        {"user_id" : id_of_user1,
         "mentioned_user_id" : id_of_user2,
         "timestamp" : ts
         },
        {"user_id" : id_of_user1,
         "mentioned_user_id" : id_of_user2,
         "timestamp" : ts
         }
    ]
}

TODO

Mettre sur des pages statiques:

• Par top 3 channel

  • Top 3 users per channel
  • Top 3 topic per channel

• Les plus actifs

  • Top 5 ecrivains (et leurs top 3 topics)
  • Top 5 channels
  • Top 6 topics

• Connected

  • Les 3 plus mentionnes
  • Les 3 qui ont le plus de mentions
  • Les 3 qui réagissent le plus
  • Les 3 auxquels on réagit le plus