Domain-specific User Research Articles

ChatReview: A ChatGPT-enabled natural language processing framework to study domain-specific user reviews

Intelligent search engines including pre-trained generative transformers (GPT) have revolutionized the user search experience. Several fields including e-commerce, education, and hospitality are increasingly exploring GPT tools to study user reviews and gain critical insights to improve their service quality. However, massive user-review data and imprecise prompt engineering lead to biased, irrelevant, and impersonal search results. In addition, exposing user data to these search engines may pose privacy issues. Motivated by these factors, we present ChatReview, a ChatGPT-enabled natural language processing (NLP) framework that effectively studies domain-specific user reviews to offer relevant and personalized search results at multiple levels of granularity. The framework accomplishes this task using four phases including data collection, tokenization, query construction, and response generation. The data collection phase involves gathering domain-specific user reviews from public and private repositories. In the tokenization phase, ChatReview applies sentiment analysis to extract keywords and categorize them into various sentiment classes. This process creates a token repository that best describes the user sentiments for a given user-review data. In the query construction phase, the framework uses the token repository and domain knowledge to construct three types of ChatGPT prompts including explicit, implicit, and creative. In the response generation phase, ChatReview pipelines these prompts into ChatGPT to generate search results at varying levels of granularity. We analyze our framework using three real-world domains including education, local restaurants, and hospitality. We assert that our framework simplifies prompt engineering for general users to produce effective results while minimizing the exposure of sensitive user data to search engines. We also present a one-of-a-kind Large Language Model (LLM) peer assessment of the ChatReview framework. Specifically, we employ Google’s Bard to objectively and qualitatively analyze the various ChatReview outputs. Our Bard-based analyses yield over 90% satisfaction, establishing ChatReview as a viable survey analysis tool.

Intent Classification by the Use of Automatically Generated Knowledge Graphs

Intent classification is an essential task for goal-oriented dialogue systems for automatically identifying customers’ goals. Although intent classification performs well in general settings, domain-specific user goals can still present a challenge for this task. To address this challenge, we automatically generate knowledge graphs for targeted data sets to capture domain-specific knowledge and leverage embeddings trained on these knowledge graphs for the intent classification task. As existing knowledge graphs might not be suitable for a targeted domain of interest, our automatic generation of knowledge graphs can extract the semantic information of any domain, which can be incorporated within the classification process. We compare our results with state-of-the-art pre-trained sentence embeddings and our evaluation of three data sets shows improvement in the intent classification task in terms of precision.

On a DSML Domain Server for Fluids Transmission Pipeline Design and Modeling

Domain specific modelling languages (DSMLs) are special purpose languages that have been designed and tailored for specific application domains. With domain specific features user of the language can construct with very familiar notations and get desired outcomes. The focus in this paper is on the functionality of a domain server in a domain specific modelling language for modelling oil and gas pipeline design. A domain server is coupled to a range of pipeline physical components by a pipeline context model, each of the components having a variety of pipeline built units, attributes and values. The domain server stores data received from at least one of the range of the pipeline physical components, the data including values associated with one or more attributes of the components. The domain server derives a model type for at least one phase from the lifecycle of the pipeline design operation for the first instance of the pipeline built units, based on analytics of information stored in the pipeline built units memory and the component attributes storage, where the model type includes a set of attributes for at least one of the range of pipeline physical components. In addition, the domain server generates an orientation for performance operation of an instance including one or more actions from the pipeline built units corresponding to at least one attribute of the set of attributes Keywords: Attribute sets, model type, built units, pipeline context model (PCM)

Adversarial Learning for Cross Domain Recommendations

Existing cross domain recommender systems typically assume homogeneous user preferences across multiple domains to capture similarities of user-item interactions and to provide cross domain recommendations accordingly. Meanwhile, the heterogeneity of user behaviors is usually not well studied and captured during the recommendation process, where users might have vastly different interests in different domains. In addition, previous models focus primarily on recommendation tasks between domain pairs, and cannot be naturally extended to serve for multiple domain recommendation applications. To address these challenges, we propose to utilize the idea of adversarial learning to intelligently incorporate global user preferences and domain-specific user preferences for providing satisfying cross domain recommendations. In particular, our proposed Adversarial Cross Domain Recommendation (ACDR) model first obtains the latent representations of global user preferences from their explicit feature information, and then transforms them into domain-specific user embeddings, where we take into account user behaviors and their heterogeneous preferences among different domains. By doing so, we address the differences among user representations in the domain-specific latent space while also preserving global user preferences, as we effectively segment the distributions of domain-specific user embeddings in the shared latent space. The convergence of our proposed model is theoretically guaranteed. The proposed ACDR model leads to significant and consistent improvements in cross domain recommendation performance over the state-of-the-art baseline models, which we demonstrate through extensive experiments on three real-world datasets. In addition, we show that the improvements are greater on those datasets that are smaller and more sparse, on those users that have fewer interaction records in the dataset, and when user interactions from more product domains are included in the cross domain recommendation model.

The development of a tool to assess medical students’ non-technical skills – The Norwegian medical students’ non-technical skills (NorMS-NTS)

Purpose New physicians need to master non-technical skills (NTS), as high levels of NTS have been shown to increase patient safety. It has also been shown that NTS can be improved through training. This study aimed to establish the necessary NTS for Norwegian medical students to create a tool for formative and summative assessments. Methods Focus group interviews were conducted with colleagues and patients of newly graduated physicians. Interviews were then analyzed using card sort methods, and the identified NTS were used to establish a framework. Focus groups commented on a prototype of an NTS assessment tool. Finally, we conducted a search of existing tools and literature. The final tool was developed based on the combined inputs. Results We created Norwegian medical students’ non-technical skills (NorMS-NTS) assessment tool containing four main categories; together comprising 13 elements and a rating scale for the NTS of the person observed. Conclusions The NorMS-NTS represents a purpose-made tool for assessing newly graduated physicians’ NTS. It is similar to existing assessment tools but based on domain-specific user perspectives obtained through focus group interviews and feedback, integrated with results from a literature search, and with consideration of existing NTS tools.

Refining Graph Representation for Cross-Domain Recommendation Based on Edge Pruning in Latent Space

In this paper, we present refining graph representation for cross-domain recommendation (CDR) based on edge pruning considering feature distribution in a latent space. Conventional graph-based CDR methods have utilized all ratings and purchase histories of user’s products. However, some items purchased by users are not related to the domain for recommendation, and this information becomes noise when making CDR. So, the proposed method introduces edge pruning into the latest graph-based CDR method to refine graph representation. To compare the item embedding features calculated in different domains, we construct a latent space and perform edge pruning through their correlations. Additionally, we introduce a state-of-the-art graph neural network into the graph construction of the proposed method that considers the interactions between users and items thereby obtaining effective embedding features in a domain. This makes it possible to consider domain-specific user preferences and estimate embedding features with high-expressive power. Furthermore, to compare the embedding features of items in the two domains, we construct their latent spaces and project them. Edge pruning is performed using the correlation of items between the two domains on the latent space. We obtain cross domain specific graph representation through edge pruning, which improves the performance by considering the relationship between both items across domains. To the best of our knowledge, no study in the CDR field focuses on eliminating unnecessary node information. We have demonstrated the effectiveness of the proposed method by comparing several graph-based state-of-the-art methods.

Towards Sentiment-Aware Multi-Modal Dialogue Policy Learning

Creation of task-oriented dialog/virtual agent (VA) capable of managing complex domain-specific user queries pertaining to multiple intents is difficult since the agent must deal with several subtasks simultaneously. Most end-to-end dialogue systems, however, only provide user semantics as inputs from texts into the learning process and neglect other useful user behaviour and information from other modalities such as images. This stresses the benefit of incorporating multi-modal inputs for eliciting user preference in the task. Also, sentiment of the user plays a significant role in achieving maximum user/customer satisfaction during the conversation. Thus, it is also important to incorporate users’ sentiments during policy learning, especially when serving user’s composite goals. For the creation of multi-modal VA aided with sentiment for conversations encompassing multi-intents, this paper introduces a new dataset, named Vis-SentiVA: Visual and Sentiment aided VA created from open-accessed conversational dataset. We present a hierarchical reinforcement learning (HRL) typically options-based VA to learn policies for serving multi-intent dialogues. Multi-modal information (texts and images) extraction to identify user’s preference is incorporated in the learning framework. A combination of task-based and sentiment-based rewards is integrated in the hierarchical value functions for the VA to be user adaptive. Empirically, we show that all these aspects induced together in the learning framework play a vital role in acquiring higher dialogue task success and increased user contentment in the process of creating composite-natured VAs. This is the first effort in integrating sentiment-aware rewards in the multi-modal HRL framework. The paper highlights that it is indeed essential to include other modes of information extraction such as images and behavioural cues of the user such as sentiment to secure greater user contentment. This also helps in improving success of composite-natured VAs serving task-oriented dialogues.

Composer-Visual Cohort Analysis of Patient Outcomes.

Visual cohort analysis utilizing electronic health record data has become an important tool in clinical assessment of patient outcomes. In this article, we introduce Composer, a visual analysis tool for orthopedic surgeons to compare changes in physical functions of a patient cohort following various spinal procedures. The goal of our project is to help researchers analyze outcomes of procedures and facilitate informed decision-making about treatment options between patient and clinician. In collaboration with orthopedic surgeons and researchers, we defined domain-specific user requirements to inform the design. We developed the tool in an iterative process with our collaborators to develop and refine functionality. With Composer, analysts can dynamically define a patient cohort using demographic information, clinical parameters, and events in patient medical histories and then analyze patient-reported outcome scores for the cohort over time, as well as compare it to other cohorts. Using Composer's current iteration, we provide a usage scenario for use of the tool in a clinical setting. We have developed a prototype cohort analysis tool to help clinicians assess patient treatment options by analyzing prior cases with similar characteristics. Although Composer was designed using patient data specific to orthopedic research, we believe the tool is generalizable to other healthcare domains. A long-term goal for Composer is to develop the application into a shared decision-making tool that allows translation of comparison and analysis from a clinician-facing interface into visual representations to communicate treatment options to patients.

Revisiting the Users Award Programme from a Value Sensitive Design Perspective

<div class="page" title="Page 1"><div class="layoutArea"><div class="column"><p><span>The goal of the UsersAward (UA) programme is to develop and maintain a strategy for enhancing the quality of workplace software through on-going user-driven quality assessment. Key activities are development of sets of quality criteria, as the U</span><span>SER </span><span>C</span><span>ERTIFIED </span><span>2002 and 2006 instruments, and performing large domain specific user satisfaction surveys building on these quality criteria. In 2005 we performed a first analysis of the values that inform the criteria and procedure making up the 2002 instrument, using the Value Sensitive Design methodology. This paper is a follow-up of that study. We report on new types of stakeholders having engaged with the UA programme and reflect on how the conceptual considerations and explicit values of the programme have shifted as a consequence. </span></p></div></div></div>

Cloud-Based Intelligent User Interface for Cloud Manufacturing: Model, Technology, and Application

Cloud manufacturing is gradually transforming the way enterprises do business from traditional production-oriented manufacturing to service-oriented manufacturing. The development of cloud manufacturing in industry practice is closely related to domain-specific user experience. The huge amount of users with diverse roles and various requirements in manufacturing industry are facing great challenges of cloud system usability problems. Thus, user interface issues play a significant role in pushing this new area forward. In this paper, we discuss the key characteristics of intelligent user interface (IUI) for cloud manufacturing, i.e., naturality, smart mobility, self-configuration, and flexible customization. Further, a cloud-plus-IUI model for cloud end-users is presented. Then we discuss the enabling technologies, i.e., automatic configuration based on virtualization, context-aware adaption and recommendation, and multimodal interaction. Finally, we present SketchPart, a sketch-based pad system prototype for searching part drawings in the cloud, to show the advantages of the proposed cloud-plus-IUI solution.

Accessible design through shared user modelling

In the proposed thesis, research examines possible exploitations of shared user modelling in order to improve the accuracy and scope of models used for personalisation within touch screen mobile devices to improve the accessibility and usability of software applications. Previous work in adaptive user interfaces has relied on local and domain specific user models, which lack in scope and detail. Shared user models can increase the accuracy and depth of data used to adapt the interfaces and user interactions.

Data-driven user simulation for automated evaluation of spoken dialog systems

This paper proposes a novel integrated dialog simulation technique for evaluating spoken dialog systems. A data-driven user simulation technique for simulating user intention and utterance is introduced. A novel user intention modeling and generating method is proposed that uses a linear-chain conditional random field, and a two-phase data-driven domain-specific user utterance simulation method and a linguistic knowledge-based ASR channel simulation method are also presented. Evaluation metrics are introduced to measure the quality of user simulation at intention and utterance. Experiments using these techniques were carried out to evaluate the performance and behavior of dialog systems designed for car navigation dialogs and a building guide robot, and it turned out that our approach was easy to set up and showed similar tendencies to real human users.

A Software Agent for Speech Abiding Systems

Problem statement: In order to bring speech into the mainstream of business process an efficient digital signal processor is necessary. The Fast Fourier Transform (FFT) and the butter fly structure symmetry will enable the harwaring easier. With the DSP and software proposed, togetherly established by means of a system, named here as “Speech Abiding System (SAS)”, a software agent, which involves the digital representation of speech signals and the use of digital processors to analyze, synthesize, or modify such signals. The proposed SAS addresses the issues in two parts. Part I: Capturing the Speaker and the Language independent error free Speech Content for speech applications processing and Part II: To accomplish the speech content as an input to the Speech User Applications/Interface (SUI). Approach: Discrete Fourier Transform (DFT) of the speech signal is the essential ingredient to evolve this SAS and Discrete-Time Fourier Transform (DTFT) links the discrete-time domain to the continuous-frequency domain. The direct computation of DFT is prohibitively expensive in terms of the required computer operations. Fortunately, a number of “fast” transforms have been developed that are mathematically equivalent to the DFT, but which require significantly a fewer computer operations for their implementation. Results: From Part-I, the SAS able to capture an error free Speech content to facilitate the speech as a good input in the main stream of business processing. Part-II provides an environment to implement the speech user applications at a primitive level. Conclusion/Recommendations: The SAS agent along with the required hardware architecture, a Finite State Automata (FSA) machine can be created to develop global oriented domain specific speech user applications easily. It will have a major impact on interoperability and disintermediation in the Information Technology Cycle (ITC) for computer program generating.

An ontology‐based competency model for workflow activity assignment policies

Purpose – Assigning business process activities to agents (human or automated) for their performance or supervision is a critical issue in business process management. Role-based approaches are commonly used to specify work assignment policies, with roles defined as collections of capabilities and privileges required to perform job functions. The purpose of this paper is to address the activity assignment problem through a competency-based approach. In this context, an ontology-based competency model is developed to assist in identifying the competencies that exist in an organization and the competencies required, by workflow activities and in performing a competency gap analysis as a prerequisite for domain-specific user development through competency-based training. Design/methodology/approach – An approach for developing a business process activity assignment policy based on an ontology-based competency model is presented. This model is also used to define domain-specific training courses that enable users meet the competency requirements of process activities. In broad terms, the approach consists of the following steps: identification of the competencies required in order to perform the various activities involved in each business process and definition of roles based on these competencies; identification of the competencies acquired in the organization and assignment of users to roles; performance of competency gap analysis to identify the missing user competencies for role playing and identification of user development needs; and development of competency-based training scenarios intended to fill the user competency gaps. Findings – An experimental implementation of the ontology-based competency model proposed in the banking domain provided a fine-grained role structure that was based on the competencies required by business process activities, and a user-to-role assignment that closely matched the competencies required for role playing, and brought forward missing user competencies that pointed to required user training needs. Originality/value – The proposed ontology-based competency model fulfils the need for a sustained work assignment approach based on user roles. To this end, roles and users are defined as collections of required and acquired competencies, respectively. A novel approach based on ontology-based competency ontologies was also developed to fill required but missing user competencies.

Model-based development of a course of action scheduling tool

This paper shows how a formal method in the form of Coloured Petri Nets (CPNs) and the supporting CPN Tools have been used in the development of the Course of Action Scheduling Tool (COAST). The aim of COAST is to support human planners in the specification and scheduling of tasks in a Course of Action. CPNs have been used to develop a formal model of the task execution framework underlying COAST. The CPN model has been extracted in executable form from CPN Tools and embedded directly into COAST, thereby automatically bridging the gap between the formal specification and its implementation. The scheduling capabilities of COAST are based on state space exploration of the embedded CPN model. Planners interact with COAST using a domain-specific graphical user interface (GUI) that hides the embedded CPN model and analysis algorithms. This means that COAST is based on a rigorous semantical model, but the use of formal methods is transparent to the users. Trials of operational planning using COAST have been conducted within the Australian Defence Force.

A new generic scheme for functional logic programming with constraints

In this paper we propose a new generic scheme CFLP?, intended as a logical and semantic framework for lazy Constraint Functional Logic Programming over a parametrically given constraint domain ?. As in the case of the well known CLP? scheme for Constraint Logic Programming, ? is assumed to provide domain specific data values and constraints. CFLP? programs are presented as sets of constrained rewrite rules that define the behavior of possibly higher order and/or non-deterministic lazy functions over ?. As a main novelty w.r.t. previous related work, we present a Constraint Rewriting Logic CRWL? which provides a declarative semantics for CFLP? programs. This logic relies on a new formalization of constraint domains and program interpretations, which allows a flexible combination of domain specific data values and user defined data constructors, as well as a functional view of constraints.

Constraint Functional Logic Programming Revisited

In this paper we propose a new generic scheme CFLP(D), intended as a logical and semantic framework for lazy Constraint Functional Logic Programming over a parametrically given constraint domain D. As in the case of the well known CLP(D) scheme for Constraint Logic Programming, D is assumed to provide domain specific data values and constraints. CFLP(D) programs are presented as sets of constrained rewrite rules that define the behaviour of possibly higher order and/or non-deterministic lazy functions over D. As the main novelty w.r.t. previous related work, we present a Constraint Rewriting Logic CRWL(D) which provides a declarative semantics for CFLP(D) programs. This logic relies on a new formalization of constraint domains and program interpretations, which allows a flexible combination of domain specific data values and user defined data constructors, as well as a functional view of constraints.

Bibliographic Access Management at Lane Medical Library: Fin de Millennium Experimentation and Bruised-Edge Innovation

SUMMARY This paper surveys four aspects of Bibliographic Management at Lane Medical Library, Stanford University Medical Center. First, a capsulized overview of the current scope and organization provides context. Second, counts of selected form/genre headings statistically present the extent and nature of databases maintained and illustrate our emphasis on form/genre. Third, descriptive summaries of selected policies and practices currently in effect illustrate how staff are attempting to improve bibliographic access and prepare for future retrieval systems. Because many of the positions taken may be controversial, discussion includes how the impact of differences is minimized in external systems. Last, selected new and/or experimental initiatives explore near future projects to further extend and enhance bibliographic control. The potential of these options derives from a more flexible integration and deployment of traditional and digital library resources focused on domain-specific user needs. A conjectural conclusion identifies the need for radical changes in the scope and structure of bibliographic control necessary to utilize rapidly evolving technologies effectively. Lane's ongoing XML MARC experiment suggests the feasibility and necessity of replacing MARC with a less arcane scheme and posits the concept of organic bibliography as the basis for a more robust bibliographic infrastructure.

Qualifying Answers According to User Needs and Preferences

This paper presents a rigorous methodology for using annotated logic programming techniques to handle user preferences and needs in answering database queries. Two alternative transformations turn a database program into a new program that returns answers to queries according to qualitative labels. The two transformations have different semantics and are each appropriate in different situations. We have modified the standard definitions of annotated logic programs to handle user needs and preferences in databases. In the formalism, the user provides a lattice of domain-independent values that define preferences and needs and a set of domain specific user constraints qualified with lattice values. After the original database and the user constraints have been transformed into a new annotated deductive database, query-answering procedures for deductive databases are used, with minor modifications, to obtain annotated answers to queries. Because preference declaration is separated from data representation and management, preferences can be easily altered without touching the database. The resulting query language allows users to ask for answers at different preference levels.