def __init__(self):
self.url = 'https://www.youtube.com/playlist?list=PL8dPuuaLjXtN0ge7yDk_UA0ldZJdhwkoV'
self.Pre = CE.Preprocessor(self.url)
self.bowSet = self.Pre._getResult()
self.Con = CE.ConceptExtraction(self.url)
self.Clu = CL.HClustering(self.url)
self.titles = self.Pre._get_videoID_titles()[1]
def __init__(self, playlistURL):
self.playlist_url = playlistURL
self.Pre = CE.Preprocessor(self.playlist_url) #module1
self.video_titles = self.Pre._get_videotitles()
self.Con = CE.ConceptExtraction(self.playlist_url) #module2
self.bowSet = self.Pre._getResult()
self.dict_set = self.Con._createDictSet(self.bowSet)
def main():
playlist_url = 'https://www.youtube.com/playlist?list=PL8dPuuaLjXtN0ge7yDk_UA0ldZJdhwkoV'
submitCode = "1503584363%7C27ffbb267ba8d3522f0aee70b23c388d"
defineConcept = DD.DefineDistance(submitCode)
makeGraph = MG.MakeGraph()
## Concept Extraction
C = CE.ConceptExtraction(playlist_url)
max_concept, max_weight = 5, 0.07
result = C._get_onlyConcepts(max_concept, max_weight)
origins = C.Pre._get_allURLs()
## Concept Mapping (concept to its Wikipedia page)
# e.g. 'inertia'(input) -> https://en.wikipedia.org/wiki/Inertia (output)
Cmap = CM.Mapping()
concept = 'inertia'
wiki_url = Cmap._mapingConcept2Wiki(concept)
## Relation Extraction
for index in range(len(origins)):
sourceName = origins[index].split("v=")[1].split("&")[0] + ".json"
print(result[index])
print(sourceName)
conceptRelation, All_degree = defineConcept.getConceptRelation(
result[index])
print(conceptRelation)
## Start Graph
graphSource = makeGraph.py2json(result[index], conceptRelation,
All_degree)
sourceLoc = os.path.join("./Web/conceptproto/play/static/play/data/" +
sourceName)
print(sourceLoc)
with open(sourceLoc, "w") as f:
f.write(graphSource)
def testGraph():
playlistURL = 'https://www.youtube.com/playlist?list=PL8dPuuaLjXtN0ge7yDk_UA0ldZJdhwkoV'
C = CE.ConceptExtraction(playlistURL)
max_concept, max_weight = 5, 0.07
makeGraph = MG.MakeGraph()
result = C._get_onlyConcepts(max_concept, max_weight)
origins = C.Pre._get_allURLs()
conceptRelation = [[0, 1, 1, 1, 2], [2, 0, 2, 1, 2], [1, 1, 0, 1, 2],
[1, 1, 1, 0, 1], [2, 2, 2, 2, 0]]
#graphSource = makeGraph.py2json(result, conceptRelation) #err
for index in range(len(origins)):
sourceName = origins[index].split("v=")[1].split("&")[0] + ".json"
print(result[index])
print(sourceName)
submitCode = "1503382656%7Ce5c72339e330f6814ae2fe97aa5c6301"
defineConcept = DD.DefineDistance(submitCode)
conceptRelation, All_degree = defineConcept.getConceptRelation(
result[index])
print(conceptRelation)
# Start graph
graphSource = makeGraph.py2json(result[index], conceptRelation,
All_degree)
sourceLoc = os.path.join("./Web/conceptproto/play/static/play/data/" +
sourceName)
print(sourceLoc)
with open(sourceLoc, "w") as f:
f.write(graphSource)
def _linkWord2Lec(self, max_concept, max_weight, bowSet):
playlistURL = "https://www.youtube.com/playlist?list=PL8dPuuaLjXtN0ge7yDk_UA0ldZJdhwkoV"
Con = CE.ConceptExtraction(playlistURL)
Pre = CE.Preprocessor(playlistURL)
final_concept_weight = Con._get_conceptWeight(bowSet, max_concept,
max_weight)
titles = Pre._get_videoID_titles()[1]
lec_title = {}
# e.g. {1:'Motion in a Straight Line', 2: 'Derivatives', 3: 'Integrals',..}
for i in range(len(titles)):
lec_title[i + 1] = titles[i]
ConceptToLec = {}
for i in range(len(final_concept_weight)):
for word, val in final_concept_weight[i]:
if word in ConceptToLec:
ConceptToLec[word].append((val, (i + 1, lec_title[i + 1])))
else:
ConceptToLec[word] = [(val, (i + 1, lec_title[i + 1]))]
return ConceptToLec
print('\n2) ids>\n', ids)
# 3)docs
print('\n3) docs[0]>\n', docs[0])
# 4)bows
print('\n4) bows[0]>\n', bows[0])
#### 2.Concept Extraction ####
'''
parameters
1) dicSet: Term-Document Dictionary(각 문서의 "Term Frequency" 계산 결과)
2) tfidf: TF-IDF 알고리즘을 계산한 최종 결과
3) *getConcept: 개념추출 결과
- 조건: 가중치(weight) 0.07 이상, 강의 별 최대 컨셉 수 5개
'''
Con = CE.ConceptExtraction(playlist_url)
lecMaxConcept, lecMaxWeight = 5, 0.07
bowSet = Pre._getResult()
dictSet = Con._createDictSet(bowSet)
tfidf = Con._runTfIdf(bowSet)
getConcept = Con._get_conceptWeight(bowSet, lecMaxConcept, lecMaxWeight)
print('\n\n2. Concept Extraction 결과..')
# 1)dicSet
print('\n1) dicSet[0]>\n', dictSet[0])
# 2)tfidf
print('\n2) tfidf[0]>\n', tfidf[0])
# 3)getConcept
print('\n1) getConcept[0]>\n', getConcept[0])