Cleanup unused files and code

app.py

@@ -10,7 +10,6 @@
 from opentapioca.languagemodel import BOWLanguageModel
 from opentapioca.tagger import Tagger
 from opentapioca.classifier import SimpleTagClassifier
-from opentapioca.goldstandard import GoldStandardDataset
 
 logger = logging.getLogger(__name__)
 logging.basicConfig(level=logging.INFO, format='%(asctime)s %(name)-12s %(levelname)-8s %(message)s')

opentapioca/wikidatagraph.py

@@ -167,53 +167,4 @@ def get_pagerank(self, qid):
         else:
             return 0.01/self.pagerank.shape[1]
 
-    def compute_similarity(self, qida, qidb, steps=3, beta=0.5, explain=False):
-        """
-        Compute the similarity between two qids
-        """
-        va = self._get_neighbour_vector(int(qida[1:]), steps, beta)
-        vb = self._get_neighbour_vector(int(qidb[1:]), steps, beta)
-        dp = va.dot(vb.transpose())
-        if explain:
-            self._print_neighbours(va)
-            self._print_neighbours(vb)
-            prod = va.multiply(vb)
-            non_zero = prod.nonzero()[1]
-            best_id = 0
-            best_score = 0
-            for idx in range(len(non_zero)):
-                score = prod[0,non_zero[idx]]
-                if score > best_score:
-                    best_score = score
-                    best_id = non_zero[idx]
-            print('https://www.wikidata.org/wiki/Q%d' % best_id)
-        return dp[0,0]
-
-    def _get_neighbour_vector(self, id, steps, beta):
-        """
-        Returns the neighbour vector after a few iterations of the matrix
-        """
-        N = self.mat.shape[0]
-        initial_v = sparse.csr_matrix(([1],([0],[id])), shape=(1,N))
-        v = initial_v
-        for k in range(steps):
-            nv = v.dot(self.mat)
-
-            # loss compensation and recurrence
-            l1norm = nv.sum()
-            if l1norm == 0:
-                return v
-            nv *= (1./l1norm)*(1-beta)
-            nv += beta*initial_v
-
-            # update
-            v = nv
-
-        return v
-
-    def _print_neighbours(self, neighbour_vector):
-        non_zero = neighbour_vector.nonzero()[1]
-        for idx in range(len(non_zero)):
-            print('\t'.join([str(neighbour_vector[0,non_zero[idx]]), 'https://www.wikidata.org/wiki/Q%d' % non_zero[idx]]))
-