suderoy/prereq-iaai-19

Inferring Concept Prerequisite Relations from Online Educational Resources (IAAI-19)

prereq-iaai-19's Issues

Custom dataset

I am struggling to build a dataset that is compatible to your requirements.

• How do you generate vocab.txt?
• How do you generate cs_preq.csv and cs_edges.csv automatically?

Environent setting

I would like to know that what environment settings that you use in python to run this coding, like which release you use in python python. And which release you use in tensorflow.

How did do you obtain the precision, accuracy, etc for Pairwise LDA?

University Course Dataset doesn't work

What I did:

cd "/content/prereq-iaai-19/datasets/University Course Dataset"
python "../../data preprocessing/preq_preparation.py"
python "../../data preprocessing/create_bow_features.py"

cd /content/prereq-iaai-19/pairwise linked lda
make
./lda est 0.01 100 settings.txt "../datasets/University Course Dataset/bow_feature.txt" "../datasets/University Course Dataset/prerequisites.link" random "../datasets/University Course Dataset/k100_a0.01"

cd /content/prereq-iaai-19/siamese network
python siamese_fc_relu.py "University Course Dataset" "../datasets/University Course Dataset/k100_a0.01" 100

And I got this:

2021-01-04 16:27:31.306618: I tensorflow/core/common_runtime/process_util.cc:115] Creating new thread pool with default inter op setting: 2. Tune using inter_op_parallelism_threads for best performance.
Network training begins.
Traceback (most recent call last):
  File "siamese_fc_relu.py", line 190, in <module>
    batch_x1, batch_x2, batch_y, batch_cs, batch_ct = data.get_next_batch(batch_train, phase='train')
  File "/content/prereq-iaai-19/siamese network/data_processing/siamese_data_train_test.py", line 206, in get_next_batch
    x1_ = np.asarray([x_.reshape((1, self.dim)) for x_ in x1_])
  File "/content/prereq-iaai-19/siamese network/data_processing/siamese_data_train_test.py", line 206, in <listcomp>
    x1_ = np.asarray([x_.reshape((1, self.dim)) for x_ in x1_])
ValueError: cannot reshape array of size 0 into shape (1,100)

It worked fined with NPTEL_MOOC dataset

InvalidArgumentError

Hi,

I was trying to run your code by using our own datasets, but there is an error occurs. It showed: InvalidArgumentError (see above for traceback): Incompatible shapes: [1,122,512] vs. [1,125,512].

Do you know what is the problem?

How to get 345 concepts out of 406?

Hi,
For the MOOC dataset, could you please explain more on how did you match the the BoW vocabulary with the 406 unique concepts to get the final 345 concepts? Thanks!

How to get vocab.txt file?

Hi,
How did you get the vocab.txt file? Is it derived from cs_preq.csv?
Thanks!