Today we describe a method to do text summarization:

def summarize(self, text):

from gensim.models import word2vec

# load the test sample

sentences = text.split('\n')

model = word2vec.Word2Vec(sentences, size=200)

model.save_word2vec_format('/tmp/vectors.bin', binary=True)

model = word2vec.Word2Vec.load_word2vec_format('/tmp/vectors.bin', binary=True)

#get selection from text based on centroids selected from k-means classification avoiding graph

from nltk.cluster.kmeans import KMeansClusterer

NUM_CLUSTERS = 3  # user can specify this or we take a default value

kclusterer = KMeansClusterer(NUM_CLUSTERS, distance=model.similarity, repeats=25)

assigned_clusters = kclusterer.cluster(model, assign_clusters=True)

pr = translate_to_words(assigned_clusters)

# render the summary

import operator

sorted_pr = sorted(pr.items(), key=operator.itemgetter(1), reverse=True)

important = [int(i[0]) for i in sorted_x][:10]

scored_sentences = {}

for sentence in sentences:

matches = set(sentence.split()).intersection(important)

score = 0

for match in matches:

score+= pr[match]

scored_sentences[sentence]=score

reordered_sentences = [ i[0] for i in sorted(scored_sentences.items(), key=operator.itemgetter(1), reverse=True)[:10] ]

ordered_sentences = [ x for x in sentences if x in reordered_sentences ]

summary = '\n'.join(ordered_sentences)

print(summary)

return ordered_sentences

#codingexercise

Given a sequence of motions in terms of unit displacements as - G - for moving forward, L for moving to the left and R for moving to the right, determine if a robot executing the sequence can stay within a circle
Solution: We keep track of the current position and the orientation in terms of N, E, W, and S. For each movement in the sequence, we update the current co-ordinate by displacing x or y but not both and by updating the orientation. If the ending position converges towards the origin, we know the robot stays within a circle.

There is one observation for this coding exercise. There is no way to be sure that a given sequence converges in a predetermined number of steps.