TLDN

Here is the source code of "TLDN: Reinforcing Web Service Classification via Combining Various Unstructured and Structured Features"

Requirement:

• python: 3.6
• gensim: 3.6.0
• numpy: 1.19.0
• pandas: 0.23.4
• lda: 1.1.0
• argparse: 1.4.0
• networkx: 2.2
• torch: 1.5.0
• matplotlib: 3.0.0

Training steps : (Run the following program in turn )

./LDA and Data Preprocessing file:

• Mashuppre_LDA_20_cate.py
• APIpre_LDA_20_cate.py

./TF-IDF file:

• TF-IDF_mashup_API_20_cate.py

./Doc2vec and Data Preprocessing file:

• Doc2API_20_cate.py
• Doc2mashup_20_cate.py

./Node2vec file:

• Node2_mashup_20_cate_generation.py
• Node2_API_20_cate_generation.py
• graph_data to edgelist_data.py
• node2vec_main1.py
• emb_data to fina_data.py

./train_test_data_generationfile:

• mashup_train_test_data_generation.py
• API_train_test_data_generation.py

TFIDF_Doc2vec_LDA_Node2vec file:

• mashup_best_param_Classification.py
• API_best_param_Classification.py
• mashup_compare_Classification.py
• API_compare_Classification.py