Environmental Scanning In Social Media And Informatics Analytics

Environmental scanning is an approach an organization uses to monitor itsexternal environment including types of strategic external information shared andacted upon within the organization. Environmental scanning helps an organizationadapt to changing external circumstance, and can provide ‘signals’ for identifyingthreats and opportunities and gaining competitive advantages. The ability to assessthe relevance of topics and related sources in information-rich environments iscritical to organization success when scanning business enviroments. With theadvent of Web2.0, the past few years have witnessed the rapid rise of social mediasystems, such as Web forums, blogs, wikis, and media-sharing websites. Socialmedia systems contain large volumes of user generated content in various forms,from plain text to rich multi-media. With the availability of vast quantities ofinformation in social media, an organization has a great need for an automatedmethodology to scan and use this information.This dissrtation follows the design science research paradigm in MIS, byaddressing issues pertaining to the design and development of an important ITartifact capable of meeting the challenges of environmental scanning in social media.The purpose of this dissertation is to provide an understanding of the social mediaon strategy planning and knowledge management; to learn whether using datamining and social media analytics can yield better assimilation of knowledgemanagement in organization. Using Information Foraging Theory (IFT) as a kerneltheory, emphasis is placed on developing techniques for analyzing textual andideational information. A rich set of features are utilized to represent textual (e.g.,style, genres, social cues etc.) and ideational (topics, sentiments, affects, etc)information. The research revolves around a core set of algorithms used for featureselection, categorization, and analysis of textual content from social media.The dissertation is arranged in five chapters. Chapter1of this dissertationprovides an introduction to the environmental scanning problem in light of thecurrent problem of gathering and using the huge volume of information available viathe social media. In Chapter2, we propose an innovation research framework basedon IFT, supporting an active organization, exploration and selection of informationthat matches the needs of decision-makers in information processing. Chapter3relates to the textual feature extraction associated with social media analytics. Weassess different textual features to gauge radical opinions using machine learningtechniques on the messages from hate group Web forums. Chapter4comprises adetailed description in the design, development, and evaluation of a system that can endorse the managerial information gathering and filtering from social media.Experiments are conducted on Web forum messages in the domain of customercomplaint identification using information retrieval and partially supervised leaningtechniques. Chapter5is concerned with the information helpfulness evaluation fromsocial media, supporting the effective use of information in environmental scanning.Drawing on the paradigm of search and experience goods from informationeconomics, we develop and test a model of customer review helpfulness.The contributions of this dissertation can be manifested into four folds. First,we design an intergrated social media intelligent system that contributes to theeffective environmental scanning. The system supports the crucial tasks of theenvironmentcal scanning process, including information gathering, informationfiltering, and information use. Second, we present a new framework for automaticacquisition of domain-specific knowledge in social media analytics. Our approachenhances machine learning algorithms with features generated from domain-specificknowledge. This knowledge is represented by ontologies that contain hundreds ofthousands of concepts, further enriched through controlled Web crawling. Third, wepropose a novel labeling heuristic to extract high-quality content from social media.Our approach can dynamically capture the characteristics of the positive class withdiverse topics. Finally, we provide a theoretical framework to understand the contextof online reviews. Our study explore the antecedents of perceived quality of onlinecustomer reviews. Our findings can increase online retailers’ understanding of therole online reviews play in the multiple stages of the consumer’s purchase decisionprocess. The results of this study can be used to develop guidelines forword-of-mouth marketing.