Dear colleague, we highly appreciate Your participation in Datathon as a mentor and we are sure that Your contribution will make for another great Datathon. Please, read the mentor instructions for the Datathon below. https://www.datasciencesociety.net/datathon/
1. Log in to the Data Chat
Before starting to use the Data Chat read the instructions to install it and register at the following link:
1.1. Enter into your channels of interest
As soon as You log in the Data Chat, You should join the channels “General”, “Announcements” and at least one of the channels of the cases for the current Datathon (please, use the search field with # as starting symbol to see the full list of open channels). These channels are going to be the main place of Your contributions during the Datathon. You should introduce yorself with few wordsin the General channel. We have closed group called “mentors-internal” which you should request to be added. If You wish You may also join some team channels or create Your own (by clicking “+” by the search box).
1.2. Participate in online conference call
For clearing up on various questions You may have and to discuss some organizational details, we are organizing conference calls (You may choose to participate in one of them). You will be notified vie e-mail for their time.
1.3. Write preliminary guidelines for the cases you will be mentoring
First and most important You should notify us about which cases you prefer to mentor on (in the Data.Chat or over e-mail). This statement does not limit You in any way to give advice on the other cases, but it is rather useful information for us as organizers. In the weeks before the event you will be given access to write in a template article in which You would state Your advice and suggestions to the participants of how to solve the case. We only ask that You write your name in front of every paragraph written by you in these articles, so that there are no mix-ups with the other menthors. These guidelines should be ready about one week before the start of the event.
Give your guidlelines into the linked case guidelines. Every mentor should write their name and chat name in the beginning of their texts, so that there are no mix-ups with the other menthors.
- Mentors’ Guidelines For Datathon Case On IOT Prediction
- Mentors’ Guidelines For Datathon Case On Economic Time Series Prediction
- Mentors’ Guidelines For Datathon Case On TelCo predictions
- Mentors’ Guidelines For Datathon Case On Predictive Maintenance
- Mentors’ Guidelines For Datathon Case On Text Mining Classification
First log-in in the website and then don’t forget to join the appropriate Data Chat groups for each case to support the participants: #mentors_internal (write @Stephen or @Rhae1 to add you there); #case_telenor_sept_2018; #case_kaufland_sept_2018; #case_telelink_sept_2018; #case_nsi_sept_2018 , #case_ontotext_sept_2018 ;
You can use Data Chat for video calls and screen sharing with other group members and participants.
1.4 Participate in the online discussions
You should provide us with a timetable of when You are going to be available online during the two days of the Datathon. We expect that You would be available for mentorship at least a total of 6 hours spread in two time slots ot three hours. Take some time to familiarise with the cases which are relevant to Your field of expertise. Also it would be nice to log in about once a day online in the Datathon chat channel in the week before the start of the Datathon, so that You would interact and introduce Yourself with the participants.
2. During the event
2.1. General directions
Be an expert. The main reason we are inviting You as a mentor is that You have shown brilliance in Your field of expertise. We expect to put Your best foot forward and push the data scientists at the Datathon to their success.
Be inspiring motivator in the darkest hour. Although the level of expertise of our participants has always been very high, we have noticed that from time to time some teams run into seemingly thick walls working under pressure and within short deadlines. You as a mentor should be the one to pull them out of their despair and lift their spirits.
Support all teams, who ask for Your advice. As there would be various cases and a lot of participants, Your mentorship would be sought by many teams. While leaving to Your discretion as to how to mentor them or to which team You would put most effort, we would ask You to turn Your attention to every request for advice.
2.2. Online participation
Following Your timetable (but preferably throughout the whole event) You should participate in the online chat (in the respective channels). You may join as many team channels as You want, You may communicate with the teams through whatever other means of communication You want.
Teams may write You in as a mentor of their solution (in the field “supervised by”) so You may see inside their article while it is still work in progress (and not accessible to the public).
As a mentor You should provide feedback to as many teams as You feel comfortable after the event. Have in mind that many of the teams would continue working on their cases and Your support would help them tremendously.
2.3. Q&A live session
Normally around saturday morning there are many questions by the participants of the Datathon. That is why we organize live Q&A session on Saturday morning in which the industry experts and the mentors provide some clarifications about the cases and about the approaches to solve them. Your participation would be very beneficial to everyone.
3. At the end of the event
3.1. General principles
The event finishes with the mentors jurying all the team solutions by voting. Generally we are going for a two stage evaluation since we expect many participants (e.g. many teams). You should jury based mainly on the content of the articles of the team. The total time for judging the semi-finals and the finals would be around 1 hour (see the schedule of the event), but You would have access to the working versions of the article solutions in advance (since Saturday night) so You would have some preliminary time to get a general look over the various teams’ solutions.
The teams will be assigned in groups of 4-5 (working on the same case, and maybe one group with various cases) on Saturday evening. Every group gets to be judged by one mentor, one domain expert (from the company which provides the case) and one DSS expert. For those 4-5 teams which You end up being judge to, You should necessary provide feedback below their article (but we highly encourage You to provide feedback to as many as possible team articles). Teams may provide and uploaded video of up to 3 minutes to explain some details. After that there will be a survey form (in your website profile) provided by DSS in which You should rank all teams within the group in ascending order of how good was their solution (rank 1 means “the best”). The team which gets ranked at the top of their group qualifies for the final.
The finalist teams would have a live 3 minute presentation (over Google hangout) followed by 2 minutes Q&A session by the Jury in the General Data.Chat channel. For this round the Jury would consist of only Mentors and DSS experts all of whom get to rank the finalists (again in a provided survey form). Again it is necessary to write feedback about the team solutions which You have not judged so far. The team with the best collective rank wins.
3.4. Voting instructions
In order to jury the teams (after the deadline):
1) go to your profile (click in top right corner of the webpage);
2) go to “Jury Articles”;
3) you should be able to see links to all the teams, you are supposed to judge;
4) read all the articles, leave feedback in the comments (this is what the participants would like, and also, this enriches the content of the solution);
5) after you read them, put ranks in front of each team, where rank 1 is the best solution (according to you), second best is rank 2 and so on;
6) you should judge the teams solutions on the methodical correctness and rigorousness, but also originality of the approach. We may have metrics’ scores of the teams (depending on the case), but you should use this information for reference only.