As robots become more prolific in modern society, advances must be made in order to ensure understanding during human robot interactions. In this thesis we implement an aspect of a candidate system, Millstream systems, that could represent natural language commands as well as generate new representations for commands based on previously generated data. This thesis presents the results of using two well known syntactic and semantic parsers to generate data and implements a method of mining the data for “production rules” that dictate how to represent an uttered sentence based on the words used. These rules are then generalized using a naive method, allowing them to be applied to a larger set of inputs. Results indicate that from a corpus of 50 imperative sentences 37could be used to generate productions rules which resulted in 187 rules. These rules could then be generalized, resulting in 147 generalized rules, a compression rate of 21.3%. Finally the entire generation process was evaluated and suggestions for extensions to the system, such as gesture recognition, are presented.
Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:umu-132653 |
Date | January 2016 |
Creators | Sutherland, Alexander |
Publisher | Umeå universitet, Institutionen för datavetenskap |
Source Sets | DiVA Archive at Upsalla University |
Language | English |
Detected Language | English |
Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
Format | application/pdf |
Rights | info:eu-repo/semantics/openAccess |
Relation | UMNAD ; 1096 |
Page generated in 0.0018 seconds