The Knowledge Graph (KG) is a model that represents structured knowledge by capturing entities and their relationships in the real world. It is widely used in search engines, recommendation systems, and natural language processing. Knowledge Representation Learning (KRL) transforms semantic information from knowledge graphs into continuous vector space representations, thereby improving knowledge acquisition and reasoning capabilities. However, current KRL faces issues such as underutilization of entity attribute information and difficulty in handling zero-shot scenarios. This paper proposes the Integrated Embedding Model (IEM) to address these issues. IEM employs a BERT-based attribute encoder and attention mechanism to weigh different attribute types, creating reliable attribute information embedding. During training, IEM merges structure and attribute representations, demonstrating excellent performance in knowledge graph completion tasks. Additionally, this paper introduces the Open Domain Representation Learning Model (ODRLM) for entity and link prediction in open domain knowledge graphs. ODRLM enhances the representation of zero-shot entities and relationships through three stages of optimization. Experimental results show that this model significantly improves both entity and relationship prediction accuracy, effectively addressing challenges in knowledge graph completion, especially in Zero-shot scenarios.
Multi-user, low-loss, and cost-efficient characteristics are highly desired for widely deployed passive optical networks (PON), which are constrained by the upstream power combining loss induced by optical splitters in optical distribution networks (ODNs). We propose a multi-user low-upstream-loss PON utilizing graded-index multi-mode fiber (GI-MMF) and a compact ODN constructed by a multi-mode transformer (MMT) for the first time. Enabled by the MMT, the ODN achieves multi-mode multiplexing for low-loss combining in upstream and power splitting in downstream, simultaneously. An implementation approach for MMT using a single MPLC is also proposed, and a 44-mode MPLC-based MMT is designed for verification. Simulation results show that the device achieves a combining loss of lower than 1.13 dB and a modal crosstalk of lower than −19.5 dB in upstream and an excess splitting loss of less than 1 dB in downstream for all modes. The bandwidth characteristics and tolerance are also investigated through simulation.
Weakly-coupled mode-division multiplexing (MDM) transmission technique over widely-deployed multimode fiber (MMF) is considered a promising approach to enhance the capacity of optical fiber communication systems. In order to be compatible with cost-efficient intensity-modulation/direct-detection (IM/DD) systems, effective mode-group demultiplexing approaches to simultaneously receive each mode group of MMF are highly desired. In this paper, we propose a scalable low-modal-crosstalk mode-group demultiplexer over MMF using multi-plane light conversion (MPLC), in which input Hermite-Gaussian (HG) modes of MMF are first converted to bridging modes that composed of HG00 modes distributed as a right-angled triangle in Cartesian coordinates, and then each HG00 mode belonging to the same mode group are respectively converted to different HGn0 modes at the same output for simultaneous detection. With the help of bridging modes, the MPLC-based mode-group demultiplexer can scale to demultiplex more mode groups with relatively few phase masks. A 5 mode-group demultiplexer is also design for demonstration, and simulation results show that the modal-crosstalk are lower than -22.26 dB for all mode groups.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.