{"id":1050,"date":"2018-02-17T12:16:41","date_gmt":"2018-02-17T17:16:41","guid":{"rendered":"http:\/\/haleyai.com\/wordpress\/?p=1050"},"modified":"2018-07-26T09:23:03","modified_gmt":"2018-07-26T14:23:03","slug":"dictionary-knowledge-acquisition","status":"publish","type":"post","link":"http:\/\/haleyai.com\/wordpress\/2018\/02\/17\/dictionary-knowledge-acquisition\/","title":{"rendered":"Dictionary Knowledge Acquisition"},"content":{"rendered":"

The following is motivated by Section 6359 of the California Sales and Use Tax<\/a>.\u00a0 It demonstrates how knowledge can be acquired from dictionary definitions:<\/p>\n

\"\"<\/p>\n

Here, we’ve taken a definition from WordNet<\/a> and prefixed it with the word followed by a colon and parsed it using the Linguist<\/a>.<\/p>\n

The lexical and syntactic ambiguities of this sentence are significant.\u00a0 On the first parse, over 400 derivations are returned.\u00a0 If the software is configured to clarify parts of speech, it then presents the following dialog:<\/p>\n

\"\"<\/a><\/p>\n

And after a couple of clicks, we help the parser understand the ‘or’ by clicking on the verb phrase shown below:<\/p>\n

\"\"<\/a><\/p>\n

As you can see, this leaves only 8 parses.\u00a0 Just to be safe, we reparse using the constraints we’ve provided and find less than 400 parses, meaning that we can be more confident that the correct parse is among those we review.<\/p>\n

With parsing configured in one of its more permissible modes, there are still 16 parses with the correct parse tree structure.\u00a0 The following clauses allow us to discriminate among those semantically:<\/p>\n

\"\"<\/a><\/p>\n

Users are encouraged not to worry about why these exist but to click on those that they find easiest to affirm or reject.\u00a0 So, skimming down the list very quickly:<\/p>\n