Approximate processing of massive continuous quantile queries over high-speed data streams

Quantile computation has many applications including data mining and financial data analysis. It has been shown that an is an element of-approximate summary can be maintained so that, given a quantile query d (phi, is an element of), the data item at rank [phi N] may be approximately obtained within the rank error precision is an element of N over all N data items in a data stream or in a sliding window. However, scalable online processing of massive continuous quantile queries with different phi and is an element of poses a new challenge because the summary is continuously updated with new arrivals of data items. In this paper, first we aim to dramatically reduce the number of distinct query results by grouping a set of different queries into a cluster so that they can be processed virtually as a single query while the precision requirements from users can be retained. Second, we aim to minimize the total query processing costs. Efficient algorithms are developed to minimize the total number of times for reprocessing clusters and to produce the minimum number of clusters, respectively. The techniques are extended to maintain near-optimal clustering when queries are registered and removed in an arbitrary fashion against whole data streams or sliding windows. In addition to theoretical analysis, our performance study indicates that the proposed techniques are indeed scalable with respect to the number of input queries as well as the number of items and the item arrival rate in a data stream.

Collaborative filtering on data streams

Collaborate Filtering is one of the most popular recommendation algorithms. Most Collaborative Filtering algorithms work with a static set of data. This paper introduces a novel approach to providing recommendations using Collaborative Filtering when user rating is received over an incoming data stream. In an incoming stream there are massive amounts of data arriving rapidly making it impossible to save all the records for later analysis. By dynamically building a decision tree for every item as data arrive, the incoming data stream is used effectively although an inevitable trade off between accuracy and amount of memory used is introduced. By adding a simple personalization step using a hierarchy of the items, it is possible to improve the predicted ratings made by each decision tree and generate recommendations in real-time. Empirical studies with the dynamically built decision trees show that the personalization step improves the overall predicted accuracy.

Matching commercial clips from TV streams using a unique, robust and compact signature

One of critical challenges in automatic recognition of TV commercials is to generate a unique, robust and compact signature. Uniqueness indicates the ability to identify the similarity among the commercial video clips which may have slight content variation. Robustness means the ability to match commercial video clips containing the same content but probably with different digitalization/encoding, some noise data, and/or transmission and recording distortion. Efficiency is about the capability of effectively matching commercial video sequences with a low computation cost and storage overhead. In this paper, we present a binary signature based method, which meets all the three criteria above, by combining the techniques of ordinal and color measurements. Experimental results on a real large commercial video database show that our novel approach delivers a significantly better performance comparing to the existing methods.

Filtering duplicate items over distributed data streams

In recent years many real time applications need to handle data streams. We consider the distributed environments in which remote data sources keep on collecting data from real world or from other data sources, and continuously push the data to a central stream processor. In these kinds of environments, significant communication is induced by the transmitting of rapid, high-volume and time-varying data streams. At the same time, the computing overhead at the central processor is also incurred. In this paper, we develop a novel filter approach, called DTFilter approach, for evaluating the windowed distinct queries in such a distributed system. DTFilter approach is based on the searching algorithm using a data structure of two height-balanced trees, and it avoids transmitting duplicate items in data streams, thus lots of network resources are saved. In addition, theoretical analysis of the time spent in performing the search, and of the amount of memory needed is provided. Extensive experiments also show that DTFilter approach owns high performance.