Recommendation of Multimedia Objects Using Metadata and Link Analysis

We investigate methods for recommending multimedia items suitable for an online multimedia sharing community and introduce a novel algorithm called UserRank for ranking multimedia items based on link analysis. We also take the initiative of applying EigenRumor from the domain of blogosphere to multimedia. Furthermore, we present a strategy for making personalized recommendation that combines UserRank with collaborative filtering. We evaluate our method with an informal user study and show that results obtained are promising.

Fast in-memory XPath search using compressed indexes

A large fraction of an XML document typically consists of text data. The XPath query language allows text search via the equal, contains, and starts-with predicates. Such predicates can be efficiently implemented using a compressed self-index of the document's text nodes. Most queries, however, contain some parts querying the text of the document, plus some parts querying the tree structure. It is therefore a challenge to choose an appropriate evaluation order for a given query, which optimally leverages the execution speeds of the text and tree indexes. Here the SXSI system is introduced. It stores the tree structure of an XML document using a bit array of opening and closing brackets plus a sequence of labels, and stores the text nodes of the document using a global compressed self-index. On top of these indexes sits an XPath query engine that is based on tree automata. The engine uses fast counting queries of the text index in order to dynamically determine whether to evaluate top-down or bottom-up with respect to the tree structure. The resulting system has several advantages over existing systems: (1) on pure tree queries (without text search) such as the XPathMark queries, the SXSI system performs on par or better than the fastest known systems MonetDB and Qizx, (2) on queries that use text search, SXSI outperforms the existing systems by 1-3 orders of magnitude (depending on the size of the result set), and (3) with respect to memory consumption, SXSI outperforms all other systems for counting-only queries.

Search for Anomalous Production of Events with Two Photons and Additional Energetic Objects at CDF

We present results of a search for anomalous production of two photons together with an electron, muon, $\tau$ lepton, missing transverse energy, or jets using $p\bar{p}$ collision data from 1.1-2.0 fb$^{-1}$ of integrated luminosity collected by the Collider Detector at Fermilab (CDF). The event yields and kinematic distributions are examined for signs of new physics without favoring a specific model of new physics. The results are consistent with the standard model expectations. The search employs several new analysis techniques that significantly reduce instrumental backgrounds in channels with an electron and missing transverse energy.

Tuning the interfacial properties of supported metal nanoclusters

Nanoclusters are objects made up of several to thousands of atoms and form a transitional state of matter between single atoms and bulk materials. Due to their large surface-to-volume ratio, nanoclusters exhibit exciting and yet poorly studied size dependent properties. When deposited directly on bare metal surfaces, the interaction of the cluster with the substrate leads to alteration of the cluster properties, making it less or even non-functional. Surfaces modified with self-assembled monolayers (SAMs) were shown to form an interesting alternative platform, because of the possibility to control wettability by decreasing the surface reactivity and to add functionalities to pre-formed nanoclusters. In this thesis, the underlying size effects and the influence of the nanocluster environment are investigated. The emphasis is on the structural and magnetic properties of nanoclusters and their interaction with thiol SAMs. We report, for the first time, a ferromagnetic-like spin-glass behaviour of uncapped nanosized Au islands tens of nanometres in size. The flattening kinetics of the nanocluster deposition on thiol SAMs are shown to be mediated mainly by the thiol terminal group, as well as the deposition energy and the particle size distribution. On the other hand, a new mechanism for the penetration of the deposited nanoclusters through the monolayers is presented, which is fundamentally different from those reported for atom deposition on alkanethiols. The impinging cluster is shown to compress the thiol layer against the Au surface and subsequently intercalate at the thiol-Au interface. The compressed thiols try then to straighten and push the cluster away from the surface. Depending on the cluster size, this restoring force may or may not enable a covalent cluster-surface bond formation, giving rise to various cluster-surface binding patterns. Compression and straightening of the thiol molecules pinpoint the elastic nature of the SAMs, which has been investigated in this thesis using nanoindentation. The nanoindenation method has been applied to SAMs of varied tail groups, giving insight into the mechanical properties of thiol modified metal surfaces.