Interacting with the User

Multidimensional data analysis is currently being discussed in terms like OLAP, data warehousing, or decision support, mainly concentrating on business applications. Numerous OLAP-tools providing flexible query facilities for datacubes are being designed and distributed. Typical analysis sessions with these kind of systems comprise long and branching sequences of exploratory analysis steps which base upon each other. While concentrating on single functions and processing steps, management of this analysis process as a whole is scarcely supported.This paper proposes a dataflow-based visual programming environment for multidimensional data analysis (VIOLA) as an approach to deal with this problem. Providing a foundation of basic operations, data processing, navigation, and user interaction, an appropriate data model (MADEIRA) is developed. Epidemiological studies, i. e. investigations of aggregate data on populations, their state of health, and potential risk factors, will serve as a leading example of a typical application area.KeywordsAggregation FunctionCategory AttributeMultidimensional DataGraph NodeAnalysis SessionThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

The authentication via accelerometer-based biometric gait recognition offers a user-friendly alternative to common authentication methods on smartphones. It has the great advantage that the authentication can be performed without user interaction. When the user is walking, his walk-pattern can be extracted from the accelerations measured using the integrated sensors of the smartphone. This pattern can be used for authentication. A study showed that users often deactivate the authentication methods of their mobile devices because they consume too much time. Because all steps necessary to perform biometric gait recognition can be executed in the background, no user interaction is necessary for the presented technique. Performing a continuous authentication while the user is walking, an up-to-date authentication result is available at any point in time. During log-in, no calculations are necessary anymore, hence there is no delay. Only in cases where the user is not walking, an alternative authentication method has to be used. This is a great benefit for the user because he has the advantages of a phone which is protected by authentication but without the disadvantages common methods impose. This high user-friendliness is likely to increase the number of smartphones for which the screen-lock is linked with an authentication. Therefore, a higher security of the data stored in smartphones can be achieved. A misuse of the stored information by an unauthorized user can be prevented. Due to the growing distribution of powerful smartphones, the number of available applications is increasing as well. These applications result in a growing amount of data stored on the devices, which make the protection of the device necessary. These data comprise e.g. addresses, appointments or GPS-information. Additionally, some applications, e.g. of e-mail-providers or social networks, require the user to authenticate himself. Often these credentials are stored by the user on the phone, such that it is not necessary to enter them each time. In case an unauthorized person has access to such a phone he can use these services without restrictions and therefore substantially harm the user. The objective of this thesis was to develop methods for accelerometer-based biometric gait recognition which achieve sufficient low error rates, as well as to demonstrate that their computational effort is low and allows for an execution on current smartphones. Because the basis of existing methods is the extraction of gait cycles (i.e. two steps) from the accelerometer data, a cycle-based method was developed and evaluated in a scenario test. This method uses raw data of the gait cycles as feature vectors and accomplishes the classification using distance functions. In addition, a further approach was selected, which does not need the time-costly and error-prone gait cycle extraction. Instead, it is using overlapping segments of a fixed time length. Several features are extracted from these segments and combined to feature vectors. Machine learning algorithms are used for classification. A benchmark of the approaches on a challenging database showed that these methods yield low equal error rates between 6% and 7% and are outperforming the cycle-based methods. These error rates were achieved under the realistic conditions that training and probe data are not collected on the same day. It was shown that five minutes of gait data are sufficient to thoroughly train the models. It should be regarded that the training data contain the different walking velocities at which the user should be recognized later on. To obtain low false rejection rates, the classification should be based on around three minutes walk data. Two of the developed methods were implemented on a smartphone. It was shown that both methods are able to perform the classification fast enough to allow for an authentication without delay for the user.

A main feature of the next generation's computing software is supporting convenient user interactions. Interactive multimedia presentations that are executed in distributed environments raise some new computer communication behaviors, which we call “interactive multimedia communications”. In this paper, we (1) classify user interaction types, (2) point out the main problems in interactive multimedia communications, (3) distinguish two approaches for solving these main problems, (4) identify the possible control architectures for processing user interactions, and (5) propose some methods and control schemes accordingly.KeywordsVideo FrameUser InteractionFlow IndexControl ArchitectureMedia UnitThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

The complexity of the cancer problem domain presents challenges not only to the medical analysis systems tasked with its analysis, but also to the users of such systems. While it is desirable to assist users in operating these medical analysis systems, prior groundwork is required before this can be achieved, such as recognising patterns in the way users create certain analyses within these systems. In this paper, we use machine learning algorithms to analyse user behaviour patterns and attempt to predict the next user interaction within the CARESS medical analysis system. Since an appropriate pre-processing scheme is essential for the performance of these algorithms, we propose the usage of a Natural Language Processing (NLP)- inspired approach to preserve some semantic cohesion of the mostly categorical features of these user interactions. Furthermore, we propose to use a sliding window that contains information about the latest user interactions in combination with Latent Dirichlet Allocation (LDA) to extract a latent topic from these last interactions and use it as additional input to the machine learning models. We compare this pre-processing scheme with other approaches that utilise one-hot encoding and feature hashing. The results of our experiments show that the sliding window LDA scheme is a promising solution, that performs better for our use case than the other evaluated pre-processing schemes. Overall, our results provide an important piece for further research and development in the area of assisting users in operating analysis systems in complex problem domains.

The design and implementation of combined mobile and touchscreen-based multimodal Web 3.0 interfaces should include new approaches of intelligent semantic mediation, knowledge acquisition and user interaction when dealing with a semantic-based digitalization of mostly unstructured textual or image-based source information. In this article, we propose a semantic-based model for those three tasks. The technical components rely on semantic web data structures in order to, first, transcend the traditional keyboard and mouse interaction metaphors, and second, provide the representation structures for more complex, collaborative interaction scenarios that may combine mobile with terminal-based interaction to accommodate the growing need to store, organize, and retrieve all these data. Interactive knowledge acquisition plays a major role in increasing the quality of automatic annotations as well as the usability of different intelligent user interfaces to control, correct, and add annotations to unstructured text and image sources. Examples are provided in the context of the Medico and Texo use cases.KeywordsKnowledge AcquisitionManual AnnotationMultimodal InteractionOntology MatchSenior RadiologistThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

At the time of this paper, the advances in light field technology already offer 3D displays that immerse the users without the need for additional viewing devices. Despite the numerous advantages and attractive capabilities of such glasses-free 3D displays, their user interface methods are quite complicated and they are currently underwhelming when compared to conventional 2D displays, due to the fact that visual feedback can only be rendered sharply on the emission surface of light field displays. The sharp rendering of user interfaces is a necessity, as blur may hinder their fundamental functions. When it comes to 2D displays, many user interaction techniques and interfaces have been devised. Rendering a user interface on a 2D display could be done in various ways, such as rendering overlays on top of the rendered scene, or by using billboards. These are extensively used in modern video games. User interaction methods have proven their importance and added efficiency to virtual environments throughout the years. Due to their overall value and usefulness, interaction techniques develop immediately as new types of displays arise. With the recent advancements in visualization technologies, user interfaces have been redesigned for use in AR, VR and MR visualization. This includes on-screen augmentation, which enables interaction with visual content on the screen. Although light field displays contain immense potentials, only basic user interfaces have been devised thus far, including FOX (Focus Sliding surface), which grants users the option to scale and to rotate 3D objects. In this paper, we visualize the theater model on real light field displays and we test the different interactions by means of a monitor room. The theatre model is analogous to real-life theatres, where viewers may observe the theatrical presentation on the stage from various angles. The motivation to choose the theater model was the fact that light field visualization similarly allows multiple simultaneous viewers within its field of view, in which the content can be observed in an angle- dependent manner. Moreover, from the users’ perspective, the theater model is thus familiar and it provides high-quality visual feedback. Furthermore, theater stages encompass a lot of interactions, including rigging and flying systems, pulleys, rotating stages, lights, curtains etc. In order to test the different interaction methods on light field displays, a theater model depicting the virtual environment was implemented. Methods for rendering the monitor room and the results of the interactions are discussed in the paper, illustrated by images of the actual visualization on light field displays. It is shown that producing plausible results with no noticeable visual artifacts is challenging, yet possible. The scientific contributions of the paper also highlight the various novel user interfaces for future light field systems and services.

In this paper, we propose a user-based video-indexing method, that automatically generates thumbnails of the most important scenes of an online video stream, by analyzing users’ interactions with a web video player. As a test bench to verify our idea we have extended the YouTube video player into the VideoSkip system. In addition, VideoSkip uses a web-database (Google Application Engine) to keep a record of some important parameters, such as the timing of basic user actions (play, pause, skip). Moreover, we implemented an algorithm that selects representative thumbnails. Finally, we populated the system with data from an experiment with nine users. We found that the VideoSkip system indexes video content by crowdsourcing implicit users interactions, such as pause and thirty seconds skip. Our early findings point toward improvements of the web video player and its thumbnail generation technique. The VideoSkip system could compliment content-based algorithms, in order to achieve efficient video-indexing in difficult videos, such as lectures or sports.KeywordsVideoIndexingThumbnailsPragmaticsSemanticscrowdsourcing

Effort has been done to optimize machine learning algorithms by applying relevant knowledges in data fields in recommendation systems. Ways are explored to discover the relationship of features independently, making the model more effective and robust. A new model, DSSMFM is proposed in this paper which combines user and item features interactions to improve the performance of recommendation systems. In this model, data are divided into user features and item features represented by one-hot vectors. The pre-training for the model is proceeded through FM, and implicit vectors are obtained for both user and item features. The implicit vectors are used as the input of DSSM, and the training of the DSSM part of the model will maximize the cosine distances of the user attributes vectors and the item attributes vectors. According to the experimental results on dataset of ICME 2019 Short Video Understanding and Recommendation Challenge, the model shows improvements on some results of the baselines.

Limitations of Smalltalk’s Model-View-Controller (MVC) paradigm are discussed. A novel approach to generalising MVC is described. Our approach based on application Entities, visual Representations and Input Dispatching (ERID) decouples the conventional interdependence of MVC’s View and Controller subclass hierarchies. Instead we have generalised the user interface capability of visible entities by allowing each entity to handle its own user interaction through the incorporation of a generic input dispatcher. The interface is easily configured by the use of a variety of generic interface controllers.KeywordsConceptual ChangeGeneric ControllerInteractive BehaviourDefault RepresentationController ClassThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

In content-based image retrieval (CBIR), a set of low-level features are extracted from an image to represent its visual content. Retrieval is performed by image example where a query image is given as input by the user and an appropriate similarity measure is used to find the best matches in the corresponding feature space. This approach suffers from the fact that there is a large discrepancy between the low-level visual features that one can extract from an image and the semantic interpretation of the image’s content that a particular user may have in a given situation. That is, users seek semantic similarity, but we can only provide similarity based on low-level visual features extracted from the raw pixel data, a situation known as the semantic gap. The selection of an appropriate similarity measure is thus an important problem. Since visual content can be represented by different attributes, the combination and importance of each set of features varies according to the user’s semantic intent. Thus, the retrieval strategy should be adaptive so that it can accommodate the preferences of different users. Relevance feedback (RF) learning has been proposed as a technique aimed at reducing the semantic gap. It works by gathering semantic information from user interaction. Based on the user’s feedback on the retrieval results, the retrieval scheme is adjusted. By providing an image similarity measure under human perception, RF learning can be seen as a form of supervised learning that finds relations between high-level semantic interpretations and low-level visual properties. That is, the feedback obtained within a single query session is used to personalize the retrieval strategy and thus enhance retrieval performance. In this chapter we present an overview of CBIR and related work on RF learning. We also present our own previous work on a RF learning-based probabilistic region relevance learning algorithm for automatically estimating the importance of each region in an image based on the user’s semantic intent.

Design of intelligent English translation teaching system combined with bidirectional encoder representation

The implementation of machine learning powered intelligent chatbots have greatly enhanced customer service and user experience in various industries. With the advancements in artificial intelligence and natural language processing, chatbots have become an essential tool for businesses to effectively communicate with their customers. Machine learning algorithms enable chatbots to analyze customer queries and provide accurate responses, mimicking human conversation. This has led to an increase in customer satisfaction as chatbots can handle large volumes of inquiries and resolve them efficiently. Moreover, chatbots are available 24/7, providing round-the-clock service to customers. Implementing chatbots in businesses has also resulted in cost savings as they reduce the need for human customer service representatives. This has allowed companies to allocate their resources to other areas of their business, ultimately leading to increased productivity. In addition to improved customer service, intelligent chatbots also enhance user experience. By continuously learning from user interactions, chatbots can personalize responses and provide relevant information, making the customer experience more seamless and efficient. Furthermore, chatbots can be integrated into various platforms such as websites, social media, and messaging applications, making them easily accessible to users. This has also led to an increase in user engagement and retention rates. The potential challenges and limitations of this technology was also discussed and recommendations provided for future research in this field. Overall, the development and implementation of machine learning powered intelligent chatbots have greatly enhanced customer service and user experience, making it a valuable tool for businesses in today's digital age.

An approach based on digital image analysis to estimate the live weights of pigs in farm environments

With the rapid advancement of Virtual Reality (VR) and Augmented Reality (AR) technologies, accurately predicting users’ gaze points in virtual environments has become crucial for improving both design quality and user interactivity. This study investigates the effectiveness of three machine learning algorithms—Random Forest, Logistic Regression, and XGBoost—using the FixationNet dataset for gaze prediction tasks. Through comprehensive experiments, Random Forest consistently demonstrates superior performance across multiple evaluation metrics, including accuracy, precision, recall, and F1-score. XGBoost follows closely behind, while Logistic Regression exhibits comparatively weaker results. Furthermore, multidimensional visualization analyses are employed to reveal deeper insights into model behavior, confirming the robustness and adaptability of Random Forest in capturing complex gaze patterns within dynamic virtual spaces. These findings provide a valuable reference for future research on user behavior modeling and immersive environment optimization.

Intelligent Virtual Environments integrate AI techniques with 3D real-time environments. As such, they can support interactive problem solving, provided the underlying AI techniques can produce solutions within a time frame matching that of user interaction. In this paper, we describe an intelligent virtual environment based on Constraint Logic Programming (CLP), integrated in a real-time 3D graphic environment. We have developed an event-based approach through which user interaction can be converted in real-time into appropriate solver queries which are then translated back into automatic reconfigurations of the Virtual Environment (VE). Additionally, this framework supports the interactive exploration of the solution space in which alternative solutions (configurations) can be found. We demonstrate the system behaviour on a configuration example. This example illustrates how solutions can be interactively refined by the user through direct manipulation of objects in the VE and how the interactive search of alternative solutions in the VE is supported by these type of systems.KeywordsVirtual RealityVirtual EnvironmentVirtual WorldUser InteractionInteractive ResolutionThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.