Graph-based Representation Model Research Articles

Resolving Data Sparsity via Aggregating Graph-Based User–App–Location Association for Location Recommendations

Personalized location recommendations aim to recommend places that users want to visit, which can save their decision-making time in daily life. However, the recommending task faces a serious data sparsity problem because users have only visited a small part of total places in a city. This problem directly leads to the difficulty in learning latent representations of users and locations. In order to tackle the data sparsity problem and make better recommendations, users’ app usage records in different locations are introduced to compensated for both users’ interests and locations’ characteristics in this paper. An attributed graph-based representation model is proposed to dig out user–app–location associations with high-order features aggregated. Extensive experiments prove that better representations of users and locations are obtained by our proposed model, thus it greatly improves location recommendation performances compared with the state-of-art methods. For example, our model achieves 13.20%, 10.1%, and 9.44% higher performance than the state-of-art (SOTA) models in Top3Hitrate, Top3Accuracy, and nDCG3, respectively, in the Telecom dataset. In the TalkingData dataset, our model achieves 9.34%, 13.35%, and 8.56% better performance than the SOTA models in Top2Hitrate, Top2Accuracy, and nDCG2, respectively. Furthermore, numerical results demonstrate that our model can effectively alleviate the data sparsity problem in recommendation systems.

A graph-based method for reconstructing entities from coordination ellipsis in medical text.

Coordination ellipsis is a linguistic phenomenon abound in medical text and is challenging for concept normalization because of difficulty in recognizing elliptical expressions referencing 2 or more entities accurately. To resolve this bottleneck, we aim to contribute a generalizable method to reconstruct concepts from medical coordinated elliptical expressions in a variety of biomedical corpora. We proposed a graph-based representation model and built a pipeline to reconstruct concepts from coordinated elliptical expressions in medical text (RECEEM). There are 4 modules: (1) identify all possible candidate conjunct pairs from original coordinated elliptical expressions, (2) calculate coefficients for candidate conjuncts using the embedding model, (3) select the most appropriate decompositions by global optimization, and (4) rebuild concepts based on a pathfinding algorithm. We evaluated the pipeline's performance on 2658 coordinated elliptical expressions from 3 different medical corpora (ie, biomedical literature, clinical narratives, and eligibility criteria from clinical trials). Precision, recall, and F1 score were calculated. The F1 scores for biomedical publications, clinical narratives, and research eligibility criteria were 0.862, 0.721, and 0.870, respectively. RECEEM outperformed 2 previously released methods. By incorporating RECEEM into 2 existing NLP tools, the F1 scores increased from 0.248 to 0.460 and from 0.287 to 0.630 on concept mapping of 1125 coordination ellipses. RECEEM improves concept normalization for medical coordinated elliptical expressions in a variety of biomedical corpora. It outperformed existing methods and significantly enhanced the performance of 2 notable NLP systems for mapping coordination ellipses in the evaluation. The algorithm is open sourced online (https://github.com/chiyuan1126/RECEEM).

Foreword: Mobility and Social Networks Confluence

Mobile devices, wearables and wireless sensor networks have been lately used tomonitor and sense real world. On one hand, processing of information that has been sensed in the real world has been used for a wide set of applications. These applications include detection of human activities, environmental and urban state monitoring, estimation of healthcare conditions, proximity and concentrations of people, epidemic trends, network and routing planning and other natural and human-generated situations describing aspects of the physical world. On the other hand, the evolution of the web in terms of user-generated content, crowd sourcing and the proliferation of online social networks have generated tremendous amount of information, describing dynamic interaction among people and their surroundings in both, digital (online) and real worlds, making people-generated information to become a new type of virtual sensor. In this respect, the amount of information available from the connected world, either belonging to virtual or to physical sources, can describe aspects of reality in a detailed manner if both sources are combined. The kind of dynamics that the integration of digital and physical worlds may generate will have an impact at different levels, involving the humansocial-environmental dimensions as well as at the network organization. In this regard, a set of novel research challenges will arise for dealing with such integration of information. This Special Issue was planned to report advances on applications, methods and approaches envisioning to address the new challenges in this field. The selected papers are roughly summarized as below: The following four papers would contribute to application aspects in the mobile and social networks confluence. The paper titled “Information Propagation through Opportunistic Communication in Mobile Social Networks” presents a theoretical framework to evaluate the performance of information propagation mobile social networks formed by short range communication technologies. The paper titled “Social and Content Hybrid Image Recommender System for Mobile Social Networks” proposes a hybrid image recommender system, which combines collaborative filtering (social techniques) with content-based techniques. The paper titled “Selective Data Deactivation Mechanism in Sustainable Area-based Cache for Mobile Social Networks” presents a resilient mechanism for the proposed area-based distributed mobile cache system under a temporary decrease of the number of terminals. The paper titled “Trigger detection using geographical relation graph for social context awareness” discusses a trigger detection mechanism based on a mixed graph-based representation model able to encode geographical and social relationships among people and social objects like stores, restaurants, and event spots. The rest of the selected papers are efforts on more fundamental issues, i.e., measurement and analysis in the mobile and social networks confluence. The paper titled “Analysis of Social Interactions through Mobile Phones” discusses the feasibility of using two parameters that can be detected through mobile phone sensing, namely interpersonal distance and relative body orientation, to provide a solid basis for inferring social interactions. The paper titled “ Urban 802.11 Community Networks for Mobile Users: Current Deployments and Prospective” analyzes the deployment and the performance of wireless networks in an urban environin order to measure the potential of currently available networks for providing Internet connectivity for mobile users. We sincerely appreciate the contributions of all the authors who made submissions to this special issue. We should also mention this special issue would not have been possible without the contributions of the numerous reviewers who O. Mayora Ibarra (*) Center for Research and Telecommunication Experimentation for Networked Communities, CREATE-NET, Via alla Cascata 56/D, Povo, 38123 Trento, Italy e-mail: oscar.mayora@create-net.org

Trigger Detection Using Geographical Relation Graph for Social Context Awareness

The concept of context awareness is believed to be a key enabler for the new ubiquitous network service paradigm brought by cloud computing platforms and smartphone OSs. In particular, autonomous context-based service customization is becoming an essential tool in this context because users cannot be expected to pick step by step the appropriate network services by manually and explicitly matching preferences for their current context. In this work, we hence focus on the core problem of how to detect changes of context for network services. In turn, detection of such changes can trigger timely system reconfigurations. We introduce a trigger detection mechanism based on a mixed graph-based representation model able to encode geographical and social relationships among people and social objects like stores, restaurants, and event spots. Our mechanism generates a trigger when a significant change in the graph takes place, and it is able to render significant changes in a geographical relationship that holds among objects socially connected with each other. The main benefits of our method are that (1) it does not require building reference models in advance, and (2) it can deal with different kinds of social objects uniformly once the graph is defined. A computer simulation scenario provides evidence on the expected performance of our method.

Integrating assembly and machining planning using graph-based representation models

Traditionally, assembly planning and machining planning are considered as two independent tasks. In assembly planning, the components to be sequenced are considered as machined and finished. In machining planning, the focus has been on machining each individual component. In previous research, machining and assembly planning are analysed separately. However, in order to achieve some design specifications, the assembly and machining operations may need to be mixed in an integrated sequence. For example, a machining operation may need to be performed on a subassembly formed by a group of components in order to complete certain geometric features. In other cases, an assembly operation cannot be performed unless certain geometric features are completely machined. Therefore, the assembly and machining operations need to be planned in a combined sequence. In this research, new graph-based representation models were developed to integrate assembly and machining planning. First, an assembly-machining operation graph was developed to represent the spatial relationships between the components as well as to express the operational precedence of the machining and assembly operations. Next, the integrated assembly and machining sequences were generated using a tree structure called the assemblymachining sequence tree. Using the graph-based methodology, all the feasible integrated assembly and machining sequences can be generated and evaluated. The main objective is to provide a complete model for integrating assembly and machining sequences. A combined evaluation can be performed to find the best sequence based on certain time and cost objectives. The presented methodology is implemented on a personal computer and several example parts are discussed.