dataset

[leleyi]

I don’t see the method to deal with a dataset, can you copy it to me.
Thank you very much

[leleyi]

I don’t see the method to deal with a dataset, can you copy it to me.
Thank you very much

How to create such a data collection.

[brochier]

Do you mean the dataset loader that is here: https://github.com/brochier/expert_finding/blob/master/expert_finding/io.py ?

[brochier]

You should get the following objects:
A_da : adjacency matrix of the document-candidate network (scipy.sparse.csr_matrix)
A_dd : adjacency matrix of the document-document network (scipy.sparse.csr_matrix)
T : raw textual content of the documents (numpy.array)
L_d : labels associated to the document (corresponding to T[L_d_mask]) (numpy.array)
L_d_mask : mask to select the labeled documents (numpy.array)
L_a : labels associated to the candidates (corresponding to A_da[:,L_d_mask]) (numpy.array)
L_a_mask : mask to select the labeled candidates (numpy.array)
tags : names of the labels of expertise (numpy.array)

By looking at https://github.com/brochier/expert_finding/blob/master/expert_finding/evaluation.py you should understand how these objects should be used. Also take a look at the README.