def calcScore(self, article):
"""
when an article is passed this function calculates its overall score, by
adding content score + feed score + feed update frequency(should be
caluculated by taking the deviation from the mean update friquency.)
"""
# Get the Article title and content in plain text.
text = purify.cleanText(article.description)
titleText = purify.cleanText(article.title)
#Calculate the Score for the texual content of the article
(textTopic, textScore) = classifier.classifyArticleText(self.topic.title, text)
textTopic = textTopic.replace("_", " ")
# print "text : %s, %s" % (textTopic, textScore)
#Calculate the Score for the title of the article.
(titleTopic,titleScore) = classifier.classifyArticleTitle(self.topic.title, titleText)
titleTopic = titleTopic.replace("_", " ")
# print "Title : %s, %s" % (titleTopic,titleScore)
#Now set the textual scores to minus values if the article "notTopic"
if textTopic ==self.topic.title:
textScore = textScore * 10000
else:
textScore = textScore * (-100)
if titleTopic ==self.topic.title+"Title":
titleScore = titleScore * 10000
else:
titleScore = titleScore * (-100)
#Get the Score for the feed from the db
scoreFeed = ScoreFeed.query.filter_by(feed = article.feed, topic = self.topic).first()
feedScore = scoreFeed.score * 100
#updateFrequencyScore - score based on the feeds update frequncy.
#less frequently updated content would get fairly better scores.
# Set weights to be given for the calculated individual scores.
#TODO: Give an option for the user to set the weights of these scores from GUI.
textScoreWeight = TEXT_SCORE_WEIGHT
titleScoreWeight = TITLE_SCORE_WEIGHT
feedScoreWeight = FEED_SCORE_WEIGHT
# updateFrequencyWeight = UPDATE_FREQUENCY_WEIGHT
finalScore = ( ( textScoreWeight * textScore ) +
( titleScoreWeight * titleScore ) +
( feedScoreWeight * feedScore ) )
# ( updateFrequencyWeight * updateFrequencyScore ) )
return finalScore
def calcScore(self, article):
"""
when an article is passed this function calculates its overall score, by
adding content score + feed score + feed update frequency(should be
caluculated by taking the deviation from the mean update friquency.)
"""
# Get the Article title and content in plain text.
text = purify.cleanText(article.description)
titleText = purify.cleanText(article.title)
#Calculate the Score for the texual content of the article
(textTopic, textScore) = classifier.classifyArticleText(self.topic.title, text)
# print "text : %s, %s" % (textTopic, textScore)
#Calculate the Score for the title of the article.
(titleTopic,titleScore) = classifier.classifyArticleTitle(self.topic.title, titleText)
# print "Title : %s, %s" % (titleTopic,titleScore)
#Now set the textual scores to minus values if the article "notTopic"
if textTopic ==self.topic.title:
textScore = textScore * 10000
else:
textScore = textScore * (-100)
if titleTopic ==self.topic.title+"Title":
titleScore = titleScore * 10000
else:
titleScore = titleScore * (-100)
#Get the Score for the feed from the db
feedScore = article.feed.numVotes * 200
#updateFrequencyScore - score based on the feeds update frequncy.
#less frequently updated content would get fairly better scores.
# Set weights to be given for the calculated individual scores.
#TODO: Give an option for the user to set the weights of these scores from GUI.
textScoreWeight = 0.55
titleScoreWeight = 0.35
feedScoreWeight = 0.1
# updateFrequencyWeight = 0.1
finalScore = ( ( textScoreWeight * textScore ) +
( titleScoreWeight * titleScore ) +
( feedScoreWeight * feedScore ) )
# ( updateFrequencyWeight * updateFrequencyScore ) )
return finalScore
