CHECK THE PRESENCE OF EACH PATTERN IN THE (1756–1791)”. THE LONGEST...

1, check the presence of each pattern in the (1756–1791)”. The longest matching substring for all 3 sentences is “Mozart sentence obtained from above for two step 1 of Algorithm 1 we specify the various instances: ways in which the question term could be i) Presence of the pattern with specified in the text. The presence of any of <ANSWER> tag matched by any these names would cause it to be tagged as the word. original question term “Mozart”. ii) Presence of the pattern in the sentence The same arrangement is also done for the with <ANSWER> tag matched by the answer term so that presence of any variant of the answer term would cause it to be treated correct answer term. exactly like the original answer term. While In our example, for the pattern “<NAME> easy to do for BIRTHDATE, this step can be was born in <ANSWER>” we check the problematic for question types such as presence of the following strings in the DEFINITION, which may contain various answer sentence i) Mozart was born in <ANY_WORD> acceptable answers. In general the input example terms have to be carefully selected ii) Mozart was born in 1756 Calculate the precision of each pattern by so that the questions they represent do not have a long list of possible answers, as this the formula P = C

a

/ C

o

where would affect the confidence of the precision C

a

= total number of patterns with the scores for each pattern. All the answers need answer term present C

o

= total number of patterns present to be enlisted to ensure a high confidence in the precision score of each pattern, in the with answer term replaced by any word present framework.