Composed image retrieval is a challenging task in the field of multi-modal learning, aiming at measuring the similarities between target images and query images with modification sentences. Most previous methods either construct feature composition for the query image and modification text or concentrate on extracting cross-modal alignments. However, these methods are prone to neglect the negative impacts of the mismatched correspondences between the hybrid-modal query and target, which could be discriminative when comparing similar instances. Besides, localized textual representations are not fully explored when learning similarities between the query and the target. To overcome the above issues, we propose a Negative-Sensitive Framework with Semantic Enhancement (NSFSE) for mining the adaptive boundaries between matched and mismatched samples with comprehensive consideration of positive and negative correspondences. It can optimize the threshold dynamically based on distributions to explore the intrinsic characteristics of positive and negative correlations, which could further facilitate accurate similarity learning. A text-guided attention mechanism after infusing cross-modal affinities on localized word features is exploited in NSFSE to explore latent semantic-related visual similarity and cross-modal similarity simultaneously. The performance of extensive experiments and comprehensive analysis on three representative datasets CIRR, FashionIQ, and Fashion200 K demonstrate the effectiveness of negative mining of similarity with semantic enhancement in the proposed NSFSE.

The Electronic Product Code (EPC) network is a collection of industrial standards designed to build an Internet of physical objects. The Object Naming Service (ONS), a directory based on the Domain Name System (DNS) is one of the important components of the EPC network. ONS provides a connection between the product code and Information Services (IS) in IoT systems. However, object identification standards used by these IoT information systems are usually customized and inhomogeneous. Therefore it is difficult to integrate these systems for real-time data sharing. To solve this problem, in this paper, we propose an extension of ONS architecture to dynamically support heterogeneous object code identification. A syntax called Formal Decoding Rule (FDR) is designed to define the rules of parsing object code, so that decoding rules can be decoupled from ONS hardcode program. Object Identifier (OID) is imported to manage heterogeneous code rules, which offers open registration of new identification standards by IoT subscribers and allows real-time updates in the ONS network. And there is a one-to-one correspondence between FDR and OID. In a typical query, ONS first detects the OID of an object, finds its corresponding FDR and then resolves the object code to corresponding resource address according to this FDR. By this mechanism, IoT applications with different object code identification standards is possible to be integrated into a larger network to implement real-time data sharing. In addition, this extended ONS architecture can dynamically support heterogeneous web service interface resolving. This paper presents experimental results to illustrate a better performance of ONS using FDR than usual ONS.

Discovery Service (DS), used to trace the movement of objects along supply chains, is an important component of the Internet of Things (IoT). DS is faced with the challenge of huge volume of data as well as large amount of parallel requests of discovery and publish. This paper proposes a new central-indexing DS system using distributed NOSQL database HBase to better support big data and parallel processing. A new storage schema of DS data is designed to optimize the discovery efficiency of DS records. The new storage schema uses object ID as row key, event timestamp as column identifier and event index content as cell value. The typical recursive discovery algorithm, which is needed in DS but often neglected in previous works, is specified to realize the full tracing of object's movement. A prototype of the DS system proposed is implemented. The experiments show that the number of concurrent discoveries the proposed DS can handle per second is about 200 times that of the DS based on RDBMS, and the number of concurrent publishes the proposed DS can process per second is about five times that of the DS based on RDBMS.

The consumer-oriented technologies categorized as social media and social networking are transforming the way that people communicate and accelerating the spread of information. Small and medium sized enterprises (SMEs) lack the capability of finance and technology, and can not afford to establish a collaborative portal for partners. With social network service increasing rapidly and being used popularly, an enterprise social collaboration platform and system architecture are proposed for improving the communication and cooperation of SMEs. Synchronization method of product catalogue which is extracted from ERP or PDM is described to share product information among partners. Social project collaboration based on social networking is analyzed, and function of social project management is designed. A social collaboration platform (www.yonyou.com) has developed by Ufida Software Company which provides enterprise profile, micro blogging, blog, group, project management, product catalog sharing, document workspace, instant message.

Large companies are looking for information and communication technology to manage their resources, and also expect partners, including suppliers, to quickly adopt appropriate technologies and business strategies. The mould industry is an important support industry during the product development process, and a link, between product designers and manufacturers. The paper presents a collaborative application portal (CAP) for mould industry to provide an application platform for collaborative product design, product data management, manufacturing information management and e-business services. A multi-tiers architecture and some pivotal technologies including of MVC development mode, information exchange, user management and security for portal, are used to construct this portal. Supporting collaborative activities between mould companies and its customers, suppliers and partners are illustrated. A portal of mould industry in Singapore, www.diemoldonline.net, is introduced as an application case.