In a typical task oriented dialogue system, the interpretation task consists of mapping from the acoustic input to a series of moves. In the simplest cases each move is just a pairing of slots and values e.g. “destination = paris”. In this paper we will describe a system for language interpretation which is designed to work with lattice based output from a class based statistical language model. The work described enables phrase and keyword spotting to be incorporated within a uniform approach which also allows information from higher-level linguistic structure to be used if it is both available and likely to be helpful.
This paper proposes a theory on information security. We argue that information security is imperfectly understood and aim to bring about an altered understanding of why efforts are made to engage in information security. The goal of information security is widely recognised as the confidentiality, integrity and availability of information however we argue that the goal is actually to simply create resources. This paper responds to calls for more theory in information systems, places the discussion in philosophical context and compares various definitions. It then identifies the key concepts of information security, describes the relationships between these concepts, as well as scope and causal explanations. The paper provides the theoretical base for understanding why information is protected, in addition to theoretical and practical implications and suggestions for future research.
In an information systems context, information privacy communication will only work if information systems meet the information needs of their users. Since the needs are neither static nor uniform, a promising approach avoiding inadequacies of ignoring differences in users’ information needs and more practical than dedicated attention to each individual user is to target information privacy communication to user archetypes. To identify such archetypes, we conduct a survey eliciting users’ information needs and apply hierarchical clustering to derive a hierarchical model of user archetypes with respect to their information privacy information needs. We identify a total of 13 archetypes on two hierarchy levels. In contrast to extant research on information privacy user archetypes focusing on information privacy attitudes, the identified information privacy user archetypes are based on information system characteristics desired by users as elicited through our survey. Thus, they yield clear input for enhancing information system design with respect to information privacy. Our research highlights differences and similarities between archetypes and enriches it with an interpretatively derived characterization of the different archetypes. The resulting archetype hierarchy serves as foundation for future research aiming to improve communication of information privacy practices.
This paper addresses the contextual dependencies related to the use of information systems security and criticizes the predominance of technical and formalized paradigm in the development and implementation of IS security policies and procedures. The underlying epistemology of our research lies in the interpretative paradigm. It explores the patterns of how the contextual use of information systems security is involved according to a business/organizational practice perspective. It elicits the detailed processes and practices that constitute the pragmatic perspective in developing information security activities.
The development of new means to attack information systems by attacking humans accessing the systems has increased the attention given to risks related to human or social aspects of information sec ...
The increased amount of attacks targeting humans accessing and using computers has made it significantly important to understand human and organizational behavior in attacks and how resilient behavior can be achieved. This paper presents a research model that attempts to understand how organizational and human factors complement each other in shaping information security behavior. The model was developed through an inductive approach, in which content domain experts were interviewed to gain a deeper understanding of the phenomena. Common patterns that were identified in the interviews were then combined with data collected through surveying the literature. Specifically, the research model includes constructs related to the organization and promotion of information security, constructs related to perceptions of information security awareness and the social conditions within an organizational setting, and individual constructs related to an individual’s perceptions of attitude, normative beliefs, and self-efficacy. Implications for continuing research and how the model will be tested empirically are discussed.
Abstract The risks associated with online transactions influencing the use of e-services and e-government services include e-privacy concerns. This study has examined the impact of e-privacy risk concerns on the acceptance of e-government services in Oman using an integrated model. The model is based on Liu, Marchewka, Lu, and Yu’s (2005) privacy-trust-behavioral intention model, the broader technology acceptance literature, and recent work on e-privacy awareness and protection. Data was collected by questionnaire from Omani citizens. The model was then tested using PLS. The study found that e-privacy risk concerns and perceptions of the protection available against risks influence citizens’ intentions to use e-government services via their influence on the perceived trustworthiness of these services. Thus trustworthiness is a factor that could be an obstacle to successful e-government services project implementation.
This paper is organised as follows: in section 2 we discuss the essential ideas behind vector-based call-routing and describe two variants of the technique experimented with here.Section 2.3 outlines how latent semantic analysis (LSA) has previously been used for information retrieval and section 2.4 gives some arguments for why a different approach to the use of LSA may be appropriate for call-routing.Section 3 describes the routing scenario used and the experiments performed, and
The hidden Markov model (HMM) has proven to be the model which made large-vocabulary automatic speech recognition (ASR) possible. The HMM is robust, versatile and has at its disposal a host of efficient algorithms to deal with training, speaker adaptation and recognition. However there is nothing uniquely speech orientated about the HMM. In fact, certain assumptions are made of speech which are known to be untrue. For example, speech is modelled as a piecewise stationary process when we know it to be continuous. Also co-articulation, which should be a rich source of information, simply provides unwanted variation. This variation is generally taken into account by modelling every phone in every context which in turn leads to problems of data sparcity, making elaborate parameter tying schemes necessary.
