Class: Ebooks::Model
- Inherits:
-
Object
- Object
- Ebooks::Model
- Defined in:
- lib/twitter_ebooks/model.rb
Instance Attribute Summary collapse
-
#hash ⇒ Object
Returns the value of attribute hash.
-
#keywords ⇒ Object
Returns the value of attribute keywords.
-
#mentions ⇒ Object
Returns the value of attribute mentions.
-
#sentences ⇒ Object
Returns the value of attribute sentences.
Class Method Summary collapse
Instance Method Summary collapse
- #consume(path) ⇒ Object
-
#find_relevant(sentences, input) ⇒ Object
Finds all relevant tokenized sentences to given input by comparing non-stopword token overlaps.
- #fix(tweet) ⇒ Object
-
#make_response(input, limit = 140, sentences = @mentions) ⇒ Object
Generates a response by looking for related sentences in the corpus and building a smaller generator from these.
- #make_statement(limit = 140, generator = nil, retry_limit = 10) ⇒ Object
- #save(path) ⇒ Object
- #valid_tweet?(tokens, limit) ⇒ Boolean
-
#verbatim?(tokens) ⇒ Boolean
Test if a sentence has been copied verbatim from original.
Instance Attribute Details
#hash ⇒ Object
Returns the value of attribute hash.
11 12 13 |
# File 'lib/twitter_ebooks/model.rb', line 11 def hash @hash end |
#keywords ⇒ Object
Returns the value of attribute keywords.
11 12 13 |
# File 'lib/twitter_ebooks/model.rb', line 11 def keywords @keywords end |
#mentions ⇒ Object
Returns the value of attribute mentions.
11 12 13 |
# File 'lib/twitter_ebooks/model.rb', line 11 def mentions @mentions end |
#sentences ⇒ Object
Returns the value of attribute sentences.
11 12 13 |
# File 'lib/twitter_ebooks/model.rb', line 11 def sentences @sentences end |
Class Method Details
.consume(txtpath) ⇒ Object
13 14 15 |
# File 'lib/twitter_ebooks/model.rb', line 13 def self.consume(txtpath) Model.new.consume(txtpath) end |
.load(path) ⇒ Object
17 18 19 |
# File 'lib/twitter_ebooks/model.rb', line 17 def self.load(path) Marshal.load(File.open(path, 'rb') { |f| f.read }) end |
Instance Method Details
#consume(path) ⇒ Object
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 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 |
# File 'lib/twitter_ebooks/model.rb', line 21 def consume(path) content = File.read(path, :encoding => 'utf-8') @hash = Digest::MD5.hexdigest(content) if path.split('.')[-1] == "json" log "Reading json corpus from #{path}" lines = JSON.parse(content, symbolize_names: true).map do |tweet| tweet[:text] end elsif path.split('.')[-1] == "csv" log "Reading CSV corpus from #{path}" content = CSV.parse(content) header = content.shift text_col = header.index('text') lines = content.map do |tweet| tweet[text_col] end else log "Reading plaintext corpus from #{path}" lines = content.split("\n") end log "Removing commented lines and sorting mentions" keeping = [] mentions = [] lines.each do |l| next if l.start_with?('#') # Remove commented lines next if l.include?('RT') || l.include?('MT') # Remove soft retweets if l.include?('@') mentions << l else keeping << l end end text = NLP.normalize(keeping.join("\n")) # Normalize weird characters mention_text = NLP.normalize(mentions.join("\n")) log "Segmenting text into sentences" statements = NLP.sentences(text) mentions = NLP.sentences(mention_text) log "Tokenizing #{statements.length} statements and #{mentions.length} mentions" @sentences = [] @mentions = [] statements.each do |s| @sentences << NLP.tokenize(s).reject do |t| t.include?('@') || t.include?('http') end end mentions.each do |s| @mentions << NLP.tokenize(s).reject do |t| t.include?('@') || t.include?('http') end end log "Ranking keywords" @keywords = NLP.keywords(@sentences) self end |
#find_relevant(sentences, input) ⇒ Object
Finds all relevant tokenized sentences to given input by comparing non-stopword token overlaps
153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 |
# File 'lib/twitter_ebooks/model.rb', line 153 def find_relevant(sentences, input) relevant = [] slightly_relevant = [] tokenized = NLP.tokenize(input).map(&:downcase) sentences.each do |sent| tokenized.each do |token| if sent.map(&:downcase).include?(token) relevant << sent unless NLP.stopword?(token) slightly_relevant << sent end end end [relevant, slightly_relevant] end |
#fix(tweet) ⇒ Object
94 95 96 97 98 99 100 101 102 103 104 105 106 |
# File 'lib/twitter_ebooks/model.rb', line 94 def fix(tweet) # This seems to require an external api call #begin # fixer = NLP.gingerice.parse(tweet) # log fixer if fixer['corrections'] # tweet = fixer['result'] #rescue Exception => e # log e.message # log e.backtrace #end NLP.htmlentities.decode tweet end |
#make_response(input, limit = 140, sentences = @mentions) ⇒ Object
Generates a response by looking for related sentences in the corpus and building a smaller generator from these
173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 |
# File 'lib/twitter_ebooks/model.rb', line 173 def make_response(input, limit=140, sentences=@mentions) # Prefer mentions relevant, slightly_relevant = find_relevant(sentences, input) if relevant.length >= 3 generator = SuffixGenerator.build(relevant) make_statement(limit, generator) elsif slightly_relevant.length >= 5 generator = SuffixGenerator.build(slightly_relevant) make_statement(limit, generator) elsif sentences.equal?(@mentions) make_response(input, limit, @sentences) else make_statement(limit) end end |
#make_statement(limit = 140, generator = nil, retry_limit = 10) ⇒ Object
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 |
# File 'lib/twitter_ebooks/model.rb', line 113 def make_statement(limit=140, generator=nil, retry_limit=10) responding = !generator.nil? generator ||= SuffixGenerator.build(@sentences) retries = 0 tweet = "" while (tokens = generator.generate(3, :bigrams)) do next if tokens.length <= 3 && !responding break if valid_tweet?(tokens, limit) retries += 1 break if retries >= retry_limit end if verbatim?(tokens) && tokens.length > 3 # We made a verbatim tweet by accident while (tokens = generator.generate(3, :unigrams)) do break if valid_tweet?(tokens, limit) && !verbatim?(tokens) retries += 1 break if retries >= retry_limit end end tweet = NLP.reconstruct(tokens) if retries >= retry_limit log "Unable to produce valid non-verbatim tweet; using \"#{tweet}\"" end fix tweet end |
#save(path) ⇒ Object
87 88 89 90 91 92 |
# File 'lib/twitter_ebooks/model.rb', line 87 def save(path) File.open(path, 'wb') do |f| f.write(Marshal.dump(self)) end self end |
#valid_tweet?(tokens, limit) ⇒ Boolean
108 109 110 111 |
# File 'lib/twitter_ebooks/model.rb', line 108 def valid_tweet?(tokens, limit) tweet = NLP.reconstruct(tokens) tweet.length <= limit && !NLP.unmatched_enclosers?(tweet) end |
#verbatim?(tokens) ⇒ Boolean
Test if a sentence has been copied verbatim from original
147 148 149 |
# File 'lib/twitter_ebooks/model.rb', line 147 def verbatim?(tokens) @sentences.include?(tokens) || @mentions.include?(tokens) end |