Opinion analysis is an area of research which deals with the computational treatment of opinion statement and subjectivity in textual data. Opinion analysis has emerged over the past couple of decades as an active area of research, as it provides solutions to the issues raised by information overload. The problem of information overload has emerged with the advancements in communication technologies which gave rise to an exponential growth in user generated subjective data available online. Opinion analysis has a rich set of applications which are used to enable opportunities for organisations such as tracking user opinions about products, social issues in communities through to engagement in political participation etc. The opinion analysis area shows hyperactivity in recent years and research at different levels of granularity has, and is being undertaken. However it is observed that there are limitations in the state-of-the-art, especially as dealing with the level of granularities on their own does not solve current research issues. Therefore a novel sentence level opinion analysis approach utilising clause and phrase level analysis is proposed. This approach uses linguistic and syntactic analysis of sentences to understand the interdependence of words within sentences, and further uses rule based analysis for phrase level analysis to calculate the opinion at each hierarchical structure of a sentence. The proposed opinion analysis approach requires lexical and contextual resources for implementation. In the context of this Thesis the approach is further presented as part of an extended unifying framework for opinion analysis resulting in the design and construction of a novel corpus. The above contributions to the field (approach, framework and corpus) are evaluated within the Thesis and are found to make improvements on existing limitations in the field, particularly with regards to opinion analysis automation. Further work is required in integrating a mechanism for greater word sense disambiguation and in lexical resource development.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:676703 |
| Date | January 2015 |
| Creators | Asmi, Amna |
| Contributors | Mundy, Darren ; Ishaya, Tanko |
| Publisher | University of Hull |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hydra.hull.ac.uk/resources/hull:12429 |
Page generated in 0.0019 seconds