riiid / ednet Goto Github PK

Since google drive service is unavailable in my place.

Thank you for your contribution to this domain, and a dataset with high quality certainly will advance the research in this field. Would it be possible for you to provide text materials (text contents, concepts, course info. etc, better in English) associated with question-ids? Because these materials are very important to develop a practical algorithm for knowledge tracing, and a dataset with such materials will greatly enhance its applicability or availability to various algorithm designs.

I am conducting research in identifying associations using Knowledge Tracing. For the research, I want to use the EdNet dataset. To build my models, however, I need to know what each integer of tag(s) and bundle_id represents. As a result, I'm writing to ask if you could give me a list of tags and bundle_ids along with their corresponding concepts or disciplines. For example, bundle_Id q2844 - Calculus or Tag 179 - Topology.

https://www.kaggle.com/c/riiid-test-answer-prediction/data?select=train.csv

it said that compressed file size is bigger than uncompressed file size.
i think it's incorrect. plz check file size

Hello,

In KT1, does the elapsed_time correpond to the prior question, as the case in the Kaggle competition, or to the current question? Also, is it the average time spent on questions from the same bundle?

Specifically, could you explain why the lag_time computed below can contain negative values:

import pandas as pd

df = pd.read_csv('u42.csv')
bundle_size = dict(df.groupby('solving_id').size())

if not df['timestamp'].is_monotonic_increasing: 
    df = df.sort_values('timestamp')

df['lag_time'] = df.apply(
    lambda r: 
        0 if r.name == 0 or r['solving_id'] == df.loc[r.name - 1, 'solving_id'] 
        else r['timestamp'] - (df.loc[r.name - bundle_size[r['solving_id'] - 1], 'timestamp'] + df.loc[r.name - 1, 'elapsed_time'] * bundle_size[r['solving_id'] - 1]), 
    axis=1
).squeeze()

print(df[df['lag_time'] < 0])

Thank you.

Hi. I found that there are many duplicate records with same user id and timestamp in KT1 and KT3. For example, for user u1, in dataset KT1, there are two records with same timestamp 1567140388553 as below:
1567140388553,219,q10649,a,57500
1567140388553,219,q10648,b,57500

Another example in KT3 for user u1 with timestamp 1567115277665:
1567115277665,respond,q4790,sprint,b,mobile
1567115277665,respond,q4790,sprint,b,mobile

The first example is very confusing because of different user responses for different questions. Moreover, it seems that it is not possible to reconstruct records in KT1 with data in KT3, due to the inconsistent timestamp recorded. I am wondering whether there are some clear issues in the dataset. Is there any way to get a cleaner version? Many thanks!

Other client can be used？