Set Of Buckets Research Articles

CREAD: A Classification-Restoration Framework with Error Adaptive Discretization for Watch Time Prediction in Video Recommender Systems

The watch time is a significant indicator of user satisfaction in video recommender systems. However, the prediction of watch time as a target variable is often hindered by its highly imbalanced distribution with a scarcity of observations for larger target values and over-populated samples for small values. State-of-the-art watch time prediction models discretize the continuous watch time into a set of buckets in order to consider the distribution of watch time. However, it is highly uninvestigated how these discrete buckets should be created from the continuous watch time distribution, and existing discretization approaches suffer from either a large learning error or a large restoration error. To address this challenge, we propose a Classification-Restoration framework with Error-Adaptive-Discretization (CREAD) to accurately predict the watch time. The proposed framework contains a discretization module, a classification module, and a restoration module. It predicts the watch time through multiple classification problems. The discretization process is a key contribution of the CREAD framework. We theoretically analyze the impacts of the discretization on the learning error and the restoration error, and then propose the error-adaptive discretization (EAD) technique to better balance the two errors, which achieves better performance over traditional discretization approaches. We conduct detailed offline evaluations on a public dataset and an industrial dataset, both showing performance gains through the proposed approach. Moreover, We have fully launched our framework to an online video platform, which resulted in a significant increase in users' video watch time by 0.29% through A/B testing. These results highlight the effectiveness of the CREAD framework in watch time prediction in video recommender systems.

Content-aware data distribution over cluster nodes

Proper data items distribution may seriously improve the performance of data processing in distributed environment. However, typical datastorage systems as well as distributed computational frameworks do not pay special attention to that aspect. In this paper author introduces two custom data items addressing methods for distributed datastorage on the example of Scalable Distributed Two-Layer Datastore. The basic idea of those methods is to preserve that data items stored on the same cluster node are similar to each other following concepts of data clustering. Still, most of the data clustering mechanisms have serious problem with data scalability which is a severe limitation in Big Data applications. The proposed methods allow to efficiently distribute data set over a set of buckets. As it was shown by the experimental results, all proposed methods generate good results efficiently in comparison to traditional clustering techniques like k-means, agglomerative and birch clustering. Distributed environment experiments shown that proper data distribution can seriously improve the effectiveness of Big Data processing.

Toys or Math Tools: Do Children’s Views of ManipulativesAffect Their Learning?

ABSTRACT This research examined whether children’s construals of mathematical manipulatives – as toys or as tools for doing mathematics – influenced their learning from a lesson with the manipulatives. Children (grades 2 and 3) were presented with a set of buckets and beanbags, and they were either given no information about the manipulatives (control) or were told that the manipulatives could be used for playing a game, for doing mathematics, or both. The manipulatives were then used in a lesson about mathematical equivalence. Participants who were instructed to view the manipulatives as math tools performed better on measures of learning, transfer, and conceptual knowledge than did participants who were not instructed to view the manipulatives in this way.

Analysis of the impact of Colour of Storage Container and Storage Conditions on Bacteria flora Water

The consequence of storage and colour of buckets on the bacteriological quality of borehole water was investigated. Water samples drawn from the borehole were collected into covered tap-fitted buckets of different colours. A set of buckets were stored outdoor while duplicates were stored indoor. The physicochemical and bacteriological parameters were analyzed. There was a gradual increase in the pH of the samples during storage. Suspended solid content, bacterial count and coliform count declined with storage. There was reduction in the total bacteria count by 94.84%; 98.72%, 91.03%; 96.15%, 92.30; 96.15 and 93.60; 97.44 from an initial 78 x 10² CFU/ml for indoor and outdoor storage in the colourless, black, green and blue buckets respectively. There was also reduction in total coliform count by 95.24%; 100.00%, 90.48%; 95.24%, 90.48%; 97.62%, 92.86%; 97.62 from an initial 42 x 10² CFU/ml for indoor and outdoor storage in the colourless, black, green and blue buckets respectively. Eleven bacteria including Bacillus subtilis and Acinetobacter haemolyticus were isolated. Proteus vulgaris and Pseudomonas aeruginosa survived till the last day of storage. Decrease in bacterial counts was prominent in the colourless and blue buckets stored outdoor making them the colour of choice for use in water storage tanks.

Efficient evaluation of query rewriting plan over materialized XML view

The query rewriting plan generation over XML views has received wide attention recently. However, little work has been done on efficient evaluation of the query rewriting plans, which is not trivial since the plan may contain an exponential size of sub-plans. This paper investigates the reason for the potentially exponential number of sub-plans, and then proposes a new space-efficient form called ABCPlan (Plan with Automata Based Combinations) to equivalently represent the original query rewriting plan. ABCPlan contains a set of buckets containing suffix paths in the query tree and an automata to indicate the combination of the suffix paths from different buckets as valid query rewriting sub-plans. We also design an evaluation method called ABCScan, which constructs a unified evaluation tree for the ABCPlan and handles the evaluation tree in one scan of the XML view. In the evaluation, we introduce node existence automata to encode the structure of the sub-tree and convert the satisfaction of the ABCPlan into the intersection problem of deterministic finite automata. The experiments show that ABCPlan based method outperforms existing methods significantly in terms of scalability and efficiency.