Leihua Qin

Web Clustering Based On Tag Set Similarity

Tagging is a service that allows users to associate a set of freely determined tags with web content. Clustering web documents with tag sets can eliminate the time-consuming preprocess of word stemming. This paper proposes a novel method to compute the similarity between tag sets and use it as the distance measure to cluster web documents into groups. Major steps in this method include computing a tag similarity matrix with set-based vector space model, smoothing the similarity matrix to obtain a set of linearly independent vectors and compute the tag set similarity based on these vectors. The experimental results show that the proposed tag set similarity measures surpasses other common similarity measures not only in the reliable derivation of clustering results, but also in clustering accuracies and efficiencies.

Content-aware Network Storage System Supporting Metadata Retrieval

Nowadays, content-based network storage has become the hot research spot of academy and corporation[1]. In order to solve the problem of hit rate decline causing by migration and achieve the content-based query, we exploit a new content-aware storage system which supports metadata retrieval to improve the query performance. Firstly, we extend the SCSI command descriptor block to enable system understand those self-defined query requests. Secondly, the extracted metadata is encoded by extensible markup language to improve the universality. Thirdly, according to the demand of information lifecycle management (ILM), we store those data in different storage level and use corresponding query strategy to retrieval them. Fourthly, as the file content identifier plays an important role in locating data and calculating block correlation, we use it to fetch files and sort query results through friendly user interface. Finally, the experiments indicate that the retrieval strategy and sort algorithm have enhanced the retrieval efficiency and precision.

Research on ATCA-based Content Aware Storage Data Gateway

This research proposes a new Storage Data Gateway (SDG) on Content Aware Storage System (CASS). The CASS consists of storage application service, content auto-tiering storage; management and control sever which is the Content Management Server (CMS). The system gives each stored object a unique content address, derived from the content itself. The physical location of stored information becomes irrelevant to users. The CMS takes advantage of the tier data architecture and decides data migration strategists. But the CMS will be overload if the system assigns the backup and data migration work to it on LAN-based model. That server does not need to control each data duplication or migration. It will be free with Sever-free model. The system needs some special device to transfer data between highspeed storage node and low-speed storage node and intermix FC storage node and IP storage node with content information instead of the responsibility of CMS like on LAN-based model. That device is a new type of SDG which needs content aware, high bandwidth, strong computing ability, and high reliability. According to those requirements, we design the Content Aware SDG (CA-SDG) based on ATCA specifications. Advanced Telecommunications Computing Architecture (ATCA) is a new series of PICMG specifications, targeted to requirements for the next generation of carrier grade communications equipment. This series of specifications incorporates the latest trends in high speed interconnect technologies, next generation processors and improved reliability, manageability and serviceability. The hardware system of CA-SDG consists of two compute units (one unit deal with commands from CMS, the other interpreter data between storage nodes), two network switch units (one is redundant), two network interface units (one group units interconnect application service layer, and the other group provide the communication of heterogeneous storage nodes), one local storage unit and one manager unit. The network switch units and compute units make up of a dual star switch network. Network interface units and local storage units which are in the form of AMC card connect Compute unit through PCI-E bus, supplying extended network interface and local storage capacity to compute unit. The manager unit is responsible for the management of whole CA-SDG device.

Using Contextual Information to Assist Desktop Search

As the volume of personal data grows, it becomes increasingly difficult to rapidly find what we need on our desktop computers. Traditional content-based desktop search tools provide a potential solution by indexing file content, while lacking of contextual information about the users activity. Context-enhanced search engines seem promising: they re-rank content-based results with the users context. However, most of them are incomplete answers: they neither consider the factor of network application (especially, Web browser), nor take full advantage of the characteristics of user behavior and file metadata on desktop computers. To address those issues, we propose a concept of ”document-window switch tree” to represent the dynamic relationship from the users context. We further collect relevant information to re-rank traditional content search results, such as reading time, frequency of window switch, file content and metadata. Search results on ranking accuracy generated by our method significantly outperform the context-enhanced search tool based on temporal locality.

Available Bandwidth Based Dynamic Load Balancing over Multiple Links

Network striping is a well-known technique that can transparently utilize parallel, redundant links in order to improve the performance and reliability of the network interface. However, most of existing scheduling schemes designed for load balancing do not consider the factor of available bandwidth of multi-link. To overcome this disadvantage, in this paper, we propose a novel method called Available Bandwidth Based Dynamic Load Balancing (ABDLB) to address this issue. Using available bandwidth obtained by periodically probing, at the host, packets are proportionally allocated to multiple links according to the probed results. The simulation experiment results demonstrate that our proposed algorithm achieves better performance in terms of remote data backup.

A Novel Computing-Enhanced Cloud Storage Model Supporting Combined Service Aware

Cloud storage has been increasing in popularity recently due to its ability to deliver virtualized storage on demand over a network. As the amount of digital resources continues to grow at an astounding rate, more and more intelligent devices (such as GPU) are embedded as computing units to enhance the performance of storage system. How to seamlessly integrate these computing resources into cloud storage is an important problem that we should confront with. In this paper, a novel computing-enhanced cloud storage model is provided, which divides coarse-grained storage service into fine-grained service modules. By combining these sub-service modules with additional computing modules, a combined service can be constructed to satisfy user’s complicated requests.

High-performance server-free backup model of content aware storage system

This paper proposes a server-free backup model deploy on Content Aware Storage System (CASS). The CASS is a new type information storage system which consists of three layers. In the middle is content management control layer which is the core layer of the storage system. Compare with host based, LAN based and LAN-Free backup model, this work design a series of models to fulfill server-free backup solution. Thos models includes backup task detective, detective agent, query backup activate condition, semantics derivation, storage resource balancer, de-duplication, copy agent. This server-free backup model not only speeds up the backup operation, but also off-loads the LAN and host computer, ensuring optimum performance and continuous data access on the network.