Concept Discovery in Youtube.com Using Factorization Method
نویسندگان
چکیده
Social media are not limited to text but also multimedia. Dailymotion, YouTube, and MySpace are examples of successful sites which allow users to share videos and interact among themselves. Due to the huge amount of videos, categorizing videos with similar contents can help users to search videos more efficiently. Unlike the traditional approach to group videos into some predefined categories, we propose to facilitate video searching with clustering from comment-based matrix factorization and to improve indexing via the generation of new concept words. Factorized component entropies are introduced for handling the difficult problem of vocabulary construction for concept discovery in social media. Since the categorization is learnt from users feedback, it can accurately represent the user sentiment on the videos. Experiments conducted by using empirical data collected from YouTube shows the effectiveness of our proposed methodologies.
منابع مشابه
$n$-factorization Property of Bilinear Mappings
In this paper, we define a new concept of factorization for a bounded bilinear mapping $f:Xtimes Yto Z$, depended on a natural number $n$ and a cardinal number $kappa$; which is called $n$-factorization property of level $kappa$. Then we study the relation between $n$-factorization property of level $kappa$ for $X^*$ with respect to $f$ and automatically boundedness and $w^*$-$w^*$-continuity...
متن کاملA new approach for building recommender system using non negative matrix factorization method
Nonnegative Matrix Factorization is a new approach to reduce data dimensions. In this method, by applying the nonnegativity of the matrix data, the matrix is decomposed into components that are more interrelated and divide the data into sections where the data in these sections have a specific relationship. In this paper, we use the nonnegative matrix factorization to decompose the user ratin...
متن کاملOn the WZ Factorization of the Real and Integer Matrices
The textit{QIF} (Quadrant Interlocking Factorization) method of Evans and Hatzopoulos solves linear equation systems using textit{WZ} factorization. The WZ factorization can be faster than the textit{LU} factorization because, it performs the simultaneous evaluation of two columns or two rows. Here, we present a method for computing the real and integer textit{WZ} and textit{ZW} factoriz...
متن کاملConcept drift detection in event logs using statistical information of variants
In recent years, business process management (BPM) has been highly regarded as an improvement in the efficiency and effectiveness of organizations. Extracting and analyzing information on business processes is an important part of this structure. But these processes are not sustainable over time and may change for a variety of reasons, such as the environment and human resources. These changes ...
متن کاملDevelopment of an Efficient Hybrid Method for Motif Discovery in DNA Sequences
This work presents a hybrid method for motif discovery in DNA sequences. The proposed method called SPSO-Lk, borrows the concept of Chebyshev polynomials and uses the stochastic local search to improve the performance of the basic PSO algorithm as a motif finder. The Chebyshev polynomial concept encourages us to use a linear combination of previously discovered velocities beyond that proposed b...
متن کامل