Module: BioBiostarsAnalytics

Defined in:
lib/bio-biostars-analytics/biostars-analytics.rb

Constant Summary collapse

@@CATEGORIES =

Categories in Biostar: Type ID Type

 1        Question
 2        Answer
 3        Comment
 4        Tutorial
 5        Blog
 6        Forum
 7        News
 8
 9        Tool
10        FixMe
11        Video
12        Job
13        Research Paper
14        Tip
15        Poll
16        Ad

16

Class Method Summary collapse

Class Method Details

.cli ⇒ Object 



305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
 
# File 'lib/bio-biostars-analytics/biostars-analytics.rb', line 305

def self.cli
  if not ARGV.length.between?(2, 3) or
     not ARGV[0].match(/\d+/) or
     not ARGV[1].match(/\d+/) or
     (ARGV.length == 3 and not ARGV[2].match(/\d+/))then
    puts 'Usage: biostars-analytics max_post_number months_look_back [min_post_number]'
    puts ''
    puts 'Required parameters:'
    puts '  max_post_number    : highest number (ID) of the post that should'
    puts '                       be mined for data; the crawler will go over'
    puts '                       posts min_post_number to max_post_number'
    puts '  months_look_back   : how many months back should queries to the'
    puts '                       Biostar API go (1 month = 30 days); default'
    puts '                       value is 1'
    puts ''
    puts 'Optional parameters:'
    puts '  min_post_number    : lowest number (ID) of the post that should'
    puts '                       be mined for data'
    puts ''
    puts 'Output (date matches the script\'s invokation):'
    puts '  <date>_crawled.tsv : data mined from crawling over posts'
    puts '  <date>_api.tsv     : data extracted from the Biostar API'
    puts ''
    puts 'Example: mining Biostars in March 2014:'
    puts '  biostars-analytics 96000 54'
    exit 1
  end

  max_post_number = ARGV[0].to_i
  months_look_back = ARGV[1].to_i
  min_post_number = 1
  min_post_number = ARGV[2].to_i if ARGV.length == 3

  # Make sure not to buffer stdout, so that it is possible to
  # snoop around whilst the script is running.
  STDOUT.sync = true

  today = Time.now.strftime('%Y%m%d')
  crawler_log = File.open("#{today}_crawled.tsv", 'w')
  api_log = File.open("#{today}_api.tsv", 'w')

  (min_post_number..max_post_number).each { |i|
      minecontent(crawler_log, i)
  }

  @post_age = {}
  @user_age = {}

  (1..months_look_back*30).to_a.reverse.each { |i|
      minehistory(api_log, i)
  }

  crawler_log.close
  api_log.close
end

.extract_date(datestring) ⇒ Object

Extract the date (day, month, year) from a Biostar forum post formatted date string.



32
33
34
35
36
37
38
39
40
 
# File 'lib/bio-biostars-analytics/biostars-analytics.rb', line 32

def self.extract_date(datestring)
    # Major headache: weird years like "3.4 years ago"
    if datestring.match(/\d+\.\d+ years ago/) then
        return Chronic.parse("#{(datestring.sub(/\d+\./, '').sub(/\s.*$/, '').to_i * 5.2).to_i} weeks ago",
                                 :now => Chronic.parse(datestring.sub(/\.\d+/, '')))
    else
        return Chronic.parse(datestring)
    end
end

.minecontent(log, id) ⇒ Object

Extracts data from the rendered forum post as well as the Biostar’s “post” API.

Algorithm:

  1. mine data from the rendered forum post

  2. retrieve limited information from Biostar’s API

  3. check that gathered data matches up

  4. log it



49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
 
# File 'lib/bio-biostars-analytics/biostars-analytics.rb', line 49

def self.minecontent(log, id)
    # This hash aggregates information about a particular Biostar question and its answers/comments:
    post = { 'id' => id }

    #
    # First: mine data from the rendered forum post
    #

    url = "http://www.biostars.org/p/#{id}/"
    page = nil

    begin
        page = open(url)
    rescue
        return
    end

    if page.base_uri.to_s != url then
        # Answer URL.
        return
    end

    # Question URL that contains the question, its answers and edits.
    doc = Hpricot(page.read)

    # Bail out if this page does not explicitly mentions a question.
    return unless doc.search('doc.title') or doc.search('doc.title')[0].inner_html.match(/^Question:/)

    users = []

    # Extract user interactions: questions asked, answered and edits being made
    times = doc.search('span.relativetime|div.lastedit').map { |element|
        element.inner_html.sub(/^[^0-9]+/, '').sub(/by\s+$/, '').split("\n").first.strip
    }
    links = (doc/'a').delete_if { |link|
        if link.get_attribute('href') then
            not link.get_attribute('href').match(/^\/u\/\d+\//) # Has to be a relative link, or we catch Dropbox link-outs too...
        else
            true
        end
    }.map { |userlink| "#{userlink.get_attribute('href').gsub(/[\/u]+/, '')}\t#{userlink.inner_html}" }
    votes = doc.search('div.vote-count').map { |vote|
        if vote.inner_html.match(/^\d+$/) then
            vote.inner_html.to_i
        else
            nil
        end
    }
    tags = doc.search('a.tag').map { |link|
        link.inner_html
    }
    # Sanity check: times and users need to match up (i.e., both arrays need to be of the same length)
    unless times.length == links.length then
        $stderr.puts "Post ##{id}: recorded times and author links do not match up (#{times.length} vs. #{links.length})."
        return
    end
    # Sanity check: there cannot be more votes than times/links
    if votes.length > times.length then
        $stderr.puts "Post ##{id}: there are more votes than recorded user actions? (#{votes.length} vs. #{links.length})"
        return
    end
    # Question/answer specific stats regarding votes:
    question_vote = votes[0]
    answer_number = votes[1..-1].compact.length
    answer_min_vote = votes[1..-1].compact.sort[0]
    answer_max_vote = votes[1..-1].compact.sort[-1]
    answer_avg_vote = nil
    answer_avg_vote = (answer_min_vote + answer_max_vote).to_f / 2.0 if answer_min_vote and answer_max_vote
    # Helper variables to deal with the "votes" array, which is shorter than the times/links arrays.
    # These variables determine when the index counter for the "votes" array is incremented and when
    # said index is valid.
    vote_used = false
    vote_index = 0
    # Go through each time occurrence/author link pair (and also consider votes):
    post['records'] = times.length
    times.each_index { |index|
        # Sanity check: first time is not an update...
        if index == 0 and times[index].match(/updated/) then
            $stderr.puts "Post ##{id}: First recorded time is also an update?"
            return
        end
        # Sanity check: first time is also not a comment...
        if index == 0 and votes[index] == nil then
            $stderr.puts "Post ##{id}: First recorded time is a comment?"
            return
        end
        action = 'answered'
        action = 'asked' if index == 0
        if votes[vote_index] == nil and not vote_used then
            action = 'commented'
            vote_used = true
        end
        if times[index].match(/updated/) then
            action = 'edited'
        else
            vote_index += 1
            vote_used = false
        end
        times[index] = times[index].sub(/^[^0-9]+/, '')
        datetime = extract_date(times[index])
        post["#{index}"] = {
            'datestring' => times[index],
            'year' => datetime.year,
            'month' => datetime.month,
            'day' => datetime.day,
            'action' => action,
            'uid' => links[index],
            'question_vote' => question_vote,
            'answer_number' => answer_number,
            'answer_min_vote' => answer_min_vote,
            'answer_max_vote' => answer_max_vote,
            'answer_avg_vote' => answer_avg_vote,
            'tags' => tags
        }
    }

    page.close

    #
    # Second: retrieve limited information from Biostar's API
    #

    url = "http://www.biostars.org/api/post/#{id}/"

    begin
        doc = JSON.parse(open(url).read)
    rescue
        return
    end
    
    # Extract the limited information the API offers:
    post['api_creation_date'] = Chronic.parse(doc['creation_date'])
    post['api_answer_number'] = doc['answer_count']
    post['api_question_vote'] = doc['score']
    post['api_type'] = doc['type']
    post['api_type_id'] = doc['type_id']

    #
    # Third: check that gathered data matches up (API and data mined results are matching)
    #

    # Warning: number of answers matches
    #
    # Cannot be used as sanity check, because the Biostar implementation actually returns
    # a wrong number of answers. For example, http://www.biostars.org/p/7542/ (20 March 2014)
    # says "4 answers" even though there are clearly just three answers being displayed.
    # The same applies to underreporting of answers, such as in http://www.biostars.org/p/10927/
    # (20 March 2014), where 12 answers are shown on the web-page, but the summary on top
    # reports only 11 answers.
    unless post['api_answer_number'] == post['0']['answer_number'] then
        $stderr.puts "Post ##{id}: number of answers differ (#{post['api_answer_number']} vs. #{post['0']['answer_number']}). Resetting number returned by API; using actual count of answers visible to the user."
        post['api_answer_number'] = post['0']['answer_number']
    end

    # Sanity check: voting score for the question matches
    unless post['api_question_vote'] == post['0']['question_vote'] then
        $stderr.puts "Post ##{id}: mismatch between API's reported question vote and data mined voting score (#{post['api_question_vote']} vs. #{post['0']['question_vote']})."
        return
    end

    #
    # Fourth: log it
    #

    (0..post['records']-1).each { |index|
        record = post["#{index}"]
        log.puts "#{post['id']}\t#{record['datestring']}\t#{record['year']}\t#{record['month']}\t#{record['day']}\t#{record['action']}\t#{record['uid']}\t#{record['question_vote']}\t#{record['answer_number']}\t#{record['answer_min_vote']}\t#{record['answer_max_vote']}\t#{record['answer_avg_vote']}\t#{record['tags'].join(',')}\t#{post['api_type']}\t#{post['api_type_id']}"
   }
end

.minehistory(log, age) ⇒ Object

Extracts data from Biostar’s “stats” API.



220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
 
# File 'lib/bio-biostars-analytics/biostars-analytics.rb', line 220

def self.minehistory(log, age)
    url = "http://www.biostars.org/api/stats/#{age}/"

    begin
        stats = JSON.parse(open(url).read)
    rescue
        return
    end
    
    # Extract the limited information the API offers:
    parseddate = Chronic.parse(stats['date'])
    stats['year'] = parseddate.year
    stats['month'] = parseddate.month
    stats['day'] = parseddate.day
    
    (1..@@CATEGORIES).each { |category|
        stats["new_posts_in_category_#{category}"] = 0
    }

    # Types of votes in Biostar:
    #   Accept
    #   Bookmark
    #   Downvote
    #   Upvote
    stats['new_votes_of_type_Accept'] = 0
    stats['new_votes_of_type_Bookmark'] = 0
    stats['new_votes_of_type_Downvote'] = 0
    stats['new_votes_of_type_Upvote'] = 0

    stats['posters'] = []
    stats['poster_ages'] = []
    stats['root_post_ages'] = []
    stats['vote_post_ages'] = []
    stats['biostarbabies'] = []

    if stats.has_key?('x_new_users') then
        stats['x_new_users'].each { |post|
            @user_age[post['id']] = age
            stats['biostarbabies'] = stats['biostarbabies'] + [ post['id'] ]
        }
        stats['new_users'] = stats['x_new_users'].length
    else
        stats['new_users'] = 0
    end

    if stats.has_key?('x_new_posts') then
        stats['x_new_posts'].each { |post|
            @post_age[post['id']] = age
            stats['posters'] = stats['posters'] + [ post['author_id'] ]
            stats['poster_ages'] = stats['poster_ages'] + [ @user_age[post['author_id']] ]
            stats['root_post_ages'] = stats['root_post_ages'] + [ @post_age[post['root_id']] ] if post['root_id'] != post['id']
            stats["new_posts_in_category_#{post['type_id']}"] = stats["new_posts_in_category_#{post['type_id']}"] + 1
        }
        stats['new_posts'] = stats['x_new_posts'].length
    else
        stats['new_posts'] = 0
    end

    # Poster age might not be applicable when having gone too far back in time...
    stats['poster_ages'].reject! { |i| i == nil }

    if stats.has_key?('x_new_votes') then
        stats['x_new_votes'].each { |vote|
            stats['vote_post_ages'] = stats['vote_post_ages'] + [ @post_age[vote['post_id']] ] if vote['type'] == 'Upvote' or vote['type'] == 'Downvote'
            stats["new_votes_of_type_#{vote['type']}"] = stats["new_votes_of_type_#{vote['type']}"] + 1
        }
        stats['new_votes'] = stats['x_new_votes'].length
    else
        stats['new_votes'] = 0
    end

    line = "#{age}\t#{stats['date']}\t#{stats['year']}\t#{stats['month']}\t#{stats['day']}\t"
    (1..@@CATEGORIES).each { |category|
        line << "#{stats["new_posts_in_category_#{category}"]}\t"
    }
    line << "#{stats['new_votes_of_type_Accept']}\t"
    line << "#{stats['new_votes_of_type_Bookmark']}\t"
    line << "#{stats['new_votes_of_type_Downvote']}\t"
    line << "#{stats['new_votes_of_type_Upvote']}\t"
    line << "#{stats['new_posts']}\t#{stats['new_votes']}\t#{stats['new_users']}\t"
    line << "#{stats['posters'].join(',')}\t#{stats['poster_ages'].join(',')}\t#{stats['root_post_ages'].join(',')}\t#{stats['vote_post_ages'].join(',')}\t#{stats['biostarbabies'].join(',')}\t"

    log.puts line
end