Natural language interfaces to databases (NLIDB) are systems that aim to bridge the gap between the languages used by humans and computers, and automatically translate natural language sentences to database queries. This thesis proposes a novel approach to NLIDB, using graph-based models. The system starts by collecting as much information as possible from existing databases and sentences, and transforms this information into a knowledge base for the system. Given a new question, the system will use this knowledge to analyze and translate the sentence into its corresponding database query statement. The graph-based NLIDB system uses English as the natural language, a relational database model, and SQL as the formal query language. In experiments performed with natural language questions ran against a large database containing information about U.S. geography, the system showed good performance compared to the state-of-the-art in the field.

This project is a graphical user interface for managing log information. Logging is an important componet of a software development cycle as well as for diagnostics of performance and monitoring of the software after deployment. The LogServer Monitor provides a graphical user interface for the display and management of logged information from a distributed environment.

In the last several years, there has been an interest in graphical user interfaces as compared with character user interfaces. The "mouse" is the device most closely associated with graphical user interfaces. Key strokes are more closely associated with character user interfaces. Given these associations, is there a preference for the keyboard or for the mouse as an input device? The determination of user preference was reduced to the determination of preference of key strokes or mouse clicks for selection of main menu items. The subjects, university students working with Microsoft Works by Microsoft Corporation, copyright 1987-1989, were learning how to use application software. While Microsoft Works was running, tracking software recorded every user key stroke and mouse click, together with data about these key strokes and mouse clicks. From the analysis of these data, common preference for the means of menu item selection was determined.

Introduction. Studies have been conducted on serendipitous encounters in the physical library environment. In this study, we aimed instead to look at features in the online library catalog that can lead to serendipitous en-counters. The aim was also to see if libraries took active steps to promote serendipity in their online library cata-logs. Method. A heuristic evaluation of four online library catalogs was conducted. The set of heuristics used was based on previous research, with the aim of identifying facets of a serendipitous environment. Semi-structured interviews were also conducted with librarians from three of the researched libraries. The results from the heuristic evaluation and the transcripts of the interviews formed the data for the analysis. Analysis. Thematic analyses were carried out on the data. The data from the heuristic evaluation and the interviews were analyzed and presented separately, leading up to a joint summary. Results. Most of the online library catalogs had features like "Similar items" and book recommendations, which can lead to serendipitous encounters. Librarians often experience a lack of influence over their online library catalogs. Librarians identified problems such as patrons expecting their online library catalog to function more like a web search engine, for example, Google Search. Conclusion. The online library catalogs have features that promote serendipitous encounters but we identified problems in how they are best designed. Using an integrated interface for the online library catalog and the library's website was one way to give the librarians more influence over their interface, which increased their likelihood to provide it with more information and content. We suggest further research with an extensive approach of serendi-pitous encounters not only in the online library catalog but also in other digital library environments such as web-sites, social media, blogs and apps. This is a two years master’s thesis in Library and Information Science.

This research tested the effect of unique combinations of interface coding and presentation techniques for the redesign of the AN/SLQ-32(V) Display Control Console (DCC). The DCC provides Navy operators with bearing and threat information for radar emitters. The task of emitter identification was used to test potential redesigns. There is no research to substantiate the current or possible redesign solutions. Thus, Experiment 1 tested potential design modifications for the DCC. The factorial combination of the following comparisons yielded eight possible design solutions: color versus monochrome coding, polar (bearing only) versus range (bearing and range) presentation, and geometric symbols versus icons. Each design was tested in three conditions of emitter density: low, medium, and high. Researchers have evaluated color and symbology with consideration to emitter display systems, but without considering how range information and emitter density effect performance. Results indicate that range information improves performance by 60%. The addition of color and the new icons also significantly improves performance (17% and 15%, respectively) as compared to the current DCC configuration (Polar, Geometric, and Monochrome). Performance was measured by time to complete a task, errors, and subjective workload. Experiment 2 considered redesign solutions not restricted to the existing hardware or software. Although discussed frequently in the interface literature, performance differences between direct-manipulation and command-key interfaces have not been validated. In this study, two interfaces were constructed to take advantage of direct manipulation and command-key interaction (DMI and CKI, respectively) styles while adding a computer-aided emitter library management system, an on-screen oscilloscope, a polygon display of emitter parameters, range information, icons, increased usage of color, and other design changes. Results indicate no differences between the CKI and DMI for the time required to perform the task or for subjective workload. Although both interfaces were designed to take advantage of their respective features, operators did not perform faster with the CKI than with the DMI. However, operators did have significantly fewer errors with the DMI than with the CKI. When compared to the existing DCC and the Range/Color/Iconic design, operators using the CKI and DMI: (1) processed twice as many emitters, (2) reduced one type of error by 50% (CKI) or 67% (DMI), (3) reduced a second type of error to zero, (4) decreased subjective workload by over 50%, and (5) maintained a higher level of performance regardless of emitter density. / Ph. D.

An exploratory study was conducted to determine if adults diagnosed as having moderate developmental disability could successfully use a personal computer and, if so, the input devices preferred and the user interface design factors to be considered when designing or selecting interactive applications for this population. Participants in the three-phase study were men and women aged 25-60 living in group homes sponsored by a social services agency. All had been professionally evaluated as moderately mentally retarded. Phase I observed reaction of participants, none of whom had ever used a computer, to two input devices (mouse and touchscreen) and a graphical user interface. Phase II built on Phase I observations to design a within-subject usability test to gather heuristic data on input devices and to develop user interface guidelines for the target population. In Phase III, a prototype was built to test the effectiveness of these guidelines. Phase I found participants successfully used a graphical user interface and most could use the mouse. In Phase II, which tested mouse, trackball, and touchscreen, the mouse was preferred, although its drag-drop times were longer. Reasons given were less fatigue and greater control of icons. The Phase III prototype consisted of two games: "Shopping," to teach money handling skills, and "Getting Dressed," to teach a basic life skill. Testing found the participants preferred screens that allowed them to control the action, that quick or unexpected screen responses were upsetting, and that strong visual feedback was important. For example, "Shopping" was redesigned to enable users to visualize purchases by dragging pictures into a shopping cart. This exploratory study found that developmentally disabled persons could use commercial computer hardware and applications effectively and were proud of their newly developed computer skills. It also derived user interface guidelines for the selection or design of applications for use with this target group. Recommendations are made for a follow-up field study to investigate the applicability of the derived guidelines to a larger population of developmentally disabled persons. / Master of Information Systems

The information technology that is available now has been developed for other populations, such as the general public, the business community, health care organizations, or the handicapped. However, much of it can be adapted to fit the needs, desires and capabilities of the elderly and computer technology can be much "friendlier" for the elderly in both a technological and a social sense. This study examines the elderly-computer technology interface. State-of -the-art computer input devices are evaluated and an ergonomic analysis is made about the interface. Finally, a new design concept is developed to solve the problems and complications arising from the aging process in terms of accessibility, safety, efficiency, ergonomic comfort of computer technology and to introduce a new and fairly unfamiliar technology to a population group. / Master of Science

No description available.

No description available.

Thesis (M. Tech.) - Central University of Technology, Free State, 2009 / The purpose of the study was to design and develop a prototype Indigenous Knowledge (IK) database that will be productive within a Knowledge Management (KM) framework specifically focused on IK. The need to develop a prototype IK database that can help standardise the work being done in the field of IK within South Africa has been established in the Indigenous Knowledge Systems (IKS) policy, which stated that “common standards would enable the integration of widely scattered and distributed references on IKS in a retrievable form. This would act as a bridge between indigenous and other knowledge systems” (IKS policy, 2004:33). In particular within the indigenous people’s organizations, holders of IK, whether individually or collectively, have a claim that their knowledge should not be exploited for elitist purposes without direct benefit to their empowerment and the improvement of their livelihoods. Establishing guidelines and a modus operandi (KM framework) are important, especially when working with communities. Researchers go into communities to gather their knowledge and never return to the communities with their results. The communities feel enraged and wronged. Creating an IK network can curb such behaviour or at least inform researchers/organisations that this behaviour is damaging. The importance of IK is that IK provides the basis for problem-solving strategies for local communities, especially the poor, which can help reduce poverty. IK is a key element of the “social capital” of the poor; their main asset to invest in the struggle for survival, to produce food, to provide shelter, or to achieve control of their own lives. It is closely intertwined with their livelihoods. Many aspects of KM and IK were discussed and a feasibility study for a KM framework was conducted to determine if any existing KM frameworks can work in an organisation that works with IK. Other factors that can influence IK are: guidelines for implementing a KM framework, information management, quality management, human factors/capital movement, leading role players in the field of IK, Intellectual Property Rights (IPR), ethics, guidelines for doing fieldwork, and a best plan for implementation. At this point, the focus changes from KM and IK to the prototype IK database and the technical design thereof. The focus is shifted to a more hands-on development by looking at the different data models and their underlying models. A well-designed database facilitates data management and becomes a valuable generator of information. A poorly designed database is likely to become a breeding ground for redundant data. The conceptual design stage used data modelling to create an abstract database structure that represents real-world objects in the most authentic way possible. The tools used to design the database are platform independent software; therefore the design can be implemented on many different platforms. An elementary prototype graphical user interface was designed in order to illustrate the database’s three main functions: adding new members, adding new IK records, and searching the IK database. The IK database design took cognisance of what is currently prevailing in South Africa and the rest of the world with respect to IK and database development. The development of the database was done in such a way as to establish a standard database design for IK systems in South Africa. The goal was to design and develop a database that can be disseminated to researchers/organisations working in the field of IK so that the use of a template database can assist work in the field. Consequently the work in the field will be collected in the same way and based on the same model. At a later stage, the databases could be interlinked and South Africa can have one large knowledge repository for IK.