def test_hin_basic_adj(): (docs, phrase_set) = generate_doc_meta_from_file(DATA_PATH + 'toy_corpus_new') toy_hin = build_hin.HIN(docs_meta=docs) print toy_hin.m_d_a print toy_hin.m_d_v print toy_hin.m_d_d
import expert_finder.bibrank as bibrank
import expert_finder.build_hin as build_hin
from test_expertfinder import generate_doc_meta_from_file, generate_phrase_topic_dist
DATA_PATH = os.path.dirname(__file__) + 'dataset/'
PHRASE_DIST_PATH = DATA_PATH + 'phrase_topic_dist/'
# the following experiments are
# conducted
k = 2
p = 27451
a = 6
v = 3
(docs, phrase_set) = generate_doc_meta_from_file(DATA_PATH + 'toy_corpus_new')
toy_hin = build_hin.HIN(p=p, a=a, v=v, docs_meta=docs)
alpha = np.ones(k) #may try different topic distribution
beta = np.ones(a) #authors - uniform
gamma = np.array([[1,1,1],
[100,1,1]])
back_phrase_prob = 1.0 / p
background_topic_dist = [back_phrase_prob] * p
fpm_topic_dist = generate_phrase_topic_dist(PHRASE_DIST_PATH + 'frequent_pattern_mining', 27451, 0.4)
phrase_dist = [background_topic_dist, fpm_topic_dist]
toy_expert_finder = ef.ExpertFinder(
K=k,
docs_meta=docs,
P=p,
