WebAngels Filter: A Violent Web Filtering Engine Using Textual and Structural Content-Based Analysis

Guermazi, Radhouane; Hammami, Mohamed; Hamadou, Abdelmajid Ben

doi:10.1007/978-3-540-70720-2_21

Cited by 4 publications

(5 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Comparing with MIL, we can find that our CMIL model makes differences in the following aspects: (1) Input bag not only contains instances' data and bag's label, but also contains a contextual graph that represents the contextual relationships among different instances in the same bag. (2) The underlying label of the instance in CMIL is a fuzzy label. (3) The label of the bag is not decided by instances' labels themselves, but depends on the contextual energy of each bag.…”

Section: B Formulation Of Cmilmentioning

confidence: 99%

“…To protect our psychological health, lots of scientific researchers investigate into web filtering and try to filter out the objectionable contents automatically. Some pornographic or violent content filters, have matured to a point where robust recognition or filter software is available [1] [2]. In contrast, the research of affective semantics of horror images is still on the stage of exploration.…”

Section: Introductionmentioning

confidence: 98%

See 1 more Smart Citation

Web Horror Image Recognition Based on Context-Aware Multi-instance Learning

Li¹,

Xiong²,

Hu³

2011

2011 IEEE 11th International Conference on Data Mining

View full text Add to dashboard Cite

Along with the ever-growing Web, horror contents sharing in the Internet has interfered with our daily life and affected our, especially children's, health. Therefore horror image recognition is becoming more important for web objectionable content filtering. This paper presents a novel context-aware multi-instance learning (CMIL) model for this task. This work is distinguished by three key contributions. Firstly, the traditional multi-instance learning is extended to context-aware multi-instance learning model through integrating an undirected graph in each bag that represents contextual relationships among instances. Secondly, by introducing a novel energy function, a heuristic optimization algorithm based on Fuzzy Support Vector Machine (FSVM) is given out to find the optimal classifier on CMIL. Finally, the CMIL is applied to recognize horror images. Experimental results on an image set collected from the Internet show that the proposed method is effective on horror image recognition.

show abstract

Section: B Formulation Of Cmilmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 98%

Web Horror Image Recognition Based on Context-Aware Multi-instance Learning

Li¹,

Xiong²,

Hu³

2011

2011 IEEE 11th International Conference on Data Mining

View full text Add to dashboard Cite

show abstract

“…Most current approaches for detecting malicious contents are based on IP, text and content [1] - [3]. IPbased approaches use a blacklist of objectionable contents [1]. However, this blacklist can only represent a limited number of malicious websites at any given moment.…”

Section: Introductionmentioning

confidence: 99%

“…In order to prevent these malicious contents, many approaches have been studied. Most current approaches for detecting malicious contents are based on IP, text and content [1] - [3]. IPbased approaches use a blacklist of objectionable contents [1].…”

Section: Introductionmentioning

confidence: 99%

“…In order to solve above problems, we employed semantic features which are components of global concept. Global concept can be divided into specific or elementary semantics that have a single concept, respectively [1]. Namely, low-level features can form semantic feature (representing a single concept), and consequently semantic features form global concept.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Malicious content filtering based on semantic features

Kim

Min

Jeon

et al. 2009

Proceedings of the 2nd International Conference on Interaction Sciences: Information Technology, Culture and Human

View full text Add to dashboard Cite

This paper proposes a method to filtering malicious contents using semantic features. In conventional content based approach, lowlevel features such as color and texture are used to filter malicious contents. But, it is difficult to detect them because of semantic gaps between the low-level features and global concepts. In this paper, global concepts are divided into several semantic features. These semantic features are used to classify the global concept of malicious contents. We design semantic features and construct semantic classifier. In experiment, we evaluate the performance to filter malicious contents by comparing low-level features and semantic features. Results show that our proposed method has better performance than the method using only low-level features.

show abstract

Horror Image Recognition Based on Emotional Attention

Xiong

et al. 2011

Computer Vision – ACCV 2010

View full text Add to dashboard Cite

WebAngels Filter: A Violent Web Filtering Engine Using Textual and Structural Content-Based Analysis

Cited by 4 publications

References 18 publications

Web Horror Image Recognition Based on Context-Aware Multi-instance Learning

Web Horror Image Recognition Based on Context-Aware Multi-instance Learning

Malicious content filtering based on semantic features

Horror Image Recognition Based on Emotional Attention

Contact Info

Product

Resources

About