Show simple item record

dc.contributor.advisorBuitelaar, Paul
dc.contributor.authorO'Riain, Seán
dc.date.accessioned2012-10-23T17:02:03Z
dc.date.available2012-10-23T17:02:03Z
dc.date.issued2012-06-18
dc.identifier.urihttp://hdl.handle.net/10379/3001
dc.description.abstractSupporting competitive business analysis of financial reports through the automated analysis and interpretation of their natural language sections, presents specific challenges including information that can be ambiguous, camouflaged, or tacitly hidden within the narrative. These sections present terminology and structural challenges for information extraction that require the application of linguistic and heuristic based domain modelling to identify the information requirement. This thesis investigates a modelling approach that incrementally builds the business analysts information requirement as a series of Semantic Paths grounded in domain linguistic and user heuristics. A Competitive Analysis Ontology (CAO) is defined to provide semantic representation of the information requirement necessary to drive linguistic analysis and information extraction. The evaluation of the CAO within the financial sub-domain of competitive analysis is investigated, through the development of the Analyst Work Bench (AWB), is presented. The AWB linguistically analyses a Form 10-Q's disclosure sections, automatically populates the CAO and provides the analyst's information requirement. The AWB leverages the CAO Semantic Paths for information search and extraction capability, to support an analyst perform a competitive analysis, with reduced manual effort. Evaluation based on design-science principles, use methods from information retrieval and information system success to determine CAO performance and usability. A controlled experiment that compares competitive analysis performance using the AWB, against its manual performed equivalent, reported a 37% performance increase using the AWB to identify relevant information. Usability evaluation further found that CAO use contributed to task structuring, and structured information provision in a manner that directly supported task performance.en_US
dc.rightsAttribution-NonCommercial-NoDerivs 3.0 Ireland
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/3.0/ie/
dc.subjectCompetative Analysisen_US
dc.subjectBusiness Intelligenceen_US
dc.subjectOntologyen_US
dc.subjectLinguistic Analysisen_US
dc.subjectLinguistic Modellingen_US
dc.titleSemantic Paths in Business Filings Analysisen_US
dc.typeThesisen_US
dc.contributor.funderScience Foundation Irelanden_US
dc.local.noteThis thesis presents an approach that assists business analysts perform a competitive business analysis of financial reports. Analyst information need is constructed as a series of Semantic Paths based on domain knowledge, than combined with linguistic analysis and information extraction, helps easier identification of relevant information. Comparison with manually conducted competitive analysis report a performance increase of 37%.en_US
dc.local.finalYesen_US
nui.item.downloads1716


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution-NonCommercial-NoDerivs 3.0 Ireland
Except where otherwise noted, this item's license is described as Attribution-NonCommercial-NoDerivs 3.0 Ireland