Text Spam Research Articles

A New Bio-Inspired Method for Spam Image-Based Emails Filtering

Electronic mail has become one of the most popular and frequently used channels for personal and professional online communication. Despite its benefits, e-mail faces a major security problem, which is the daily reception of a large number of unsolicited electronic messages, known as “spam emails.” Today, most electronic mail systems have simple spam filtering mechanisms based on text spam filtering technologies. To circumvent these filters, spammers are introducing new techniques of embedding spam messages in the image attached to the mail. In this article, the authors propose a new method for spam image filtering. The proposed system can distinguish between legitimate images from spam images based on the texture characteristics of the image attached to an email. From each image, around 20 characteristics can be extracted from the gray level co-occurrence matrix (GLCM). Then, to classify the images as spam or ham, the authors use a new metaheuristic nature-inspired model for building classifiers based on the social worker bees and enhanced nearest-centroid classification method.

Improvements for research data repositories: The case of text spam

Current research has evolved in such a way scientists must not only adequately describe the algorithms they introduce and the results of their application, but also ensure the possibility of reproducing the results and comparing them with those obtained through other approximations. In this context, public data sets (sometimes shared through repositories) are one of the most important elements for the development of experimental protocols and test benches. This study has analysed a significant number of CS/ML ( Computer Science/ Machine Learning) research data repositories and data sets and detected some limitations that hamper their utility. Particularly, we identify and discuss the following demanding functionalities for repositories: (1) building customised data sets for specific research tasks, (2) facilitating the comparison of different techniques using dissimilar pre-processing methods, (3) ensuring the availability of software applications to reproduce the pre-processing steps without using the repository functionalities and (4) providing protection mechanisms for licencing issues and user rights. To show the introduced functionality, we created STRep (Spam Text Repository) web application which implements our recommendations adapted to the field of spam text repositories. In addition, we launched an instance of STRep in the URL https://rdata.4spam.group to facilitate understanding of this study.

Classification of Social Media Text Spam Using VAE-CNN and LSTM Model

Presently a day's human relations are kept up by online life systems. Customary connections now days are outdated. To keep up in affiliation, sharing thoughts, trade information between we utilize web-based social networking organizing locales. Web based life organizing locales like Twitter, Facebook, LinkedIn and so forth are accessible in the correspondence condition. Through Twitter media clients share their sentiments, interests, information to others by messages. Simultaneously a portion of the client's mislead the certifiable clients. These certified clients are additionally called requested clients and the clients what misguidance's identity is called spammers. These spammers present undesirable data on the non-spam clients. The non-spammers may retweet them to other people and they follow the spammers. Generally most of the spam messages are in the form of text, images and different multimedia formats. Considering all different formats in one process may not give the best classification results. In this paper address the process and classification of text spam messages. Classification of text messages is a complex task in order to achieve this deep learning based hybrid VAE-CNN and LSTM model is proposed and evaluated the model using the performance metrics of precision, recall and F measure metrics.

An unsupervised approach to detect review spam using duplicates of images, videos and Chinese texts

Intuitively, image- or video-based recommendations seem to be more reliable than those containing plain text, and these types of recommendations have recently become widely encouraged and commonly seen across opinion sharing platforms. Considering their potential for manipulation, graphs (e.g., images and videos) are more vulnerable to spam than scripts. However, most state-of-the-art solutions for opinion spam detection are exclusively devoted to natural language parsing, and less work has been done concerning photos or videos. After investigating the top two business-to-customer websites, i.e., JD.com and TMALL.com, we propose an unsupervised approach to label suspected spam based on different types of duplication across images, videos and Chinese texts. Experiments verified the effectiveness of this approach and obtained several conclusions: 1) the situation of image spam is more severe than that of video and text spam; 2) for manipulation, borrowing something from a marketing page is less attractive than stealing from other reviewers; 3) in addition to using identical texts, spammers also use fictitious rare incidents to influence customers; and 4) overlapping duplications of images, videos and texts are common.

Performance Analysis of Ml Techniques for Spam Filtering

The rise in the volume of unwanted spam emails has made the development of a lot more necessary more reliable and robust filters for antispam. Current machine learning approaches are used to excel Spam emails can be detected and filtered. Filtering solutions to text spam. The analysis discusses core principles, actions, efficacy, and Spam filtering trend for research. The first topic in the research study aims at the requests Machine learning approaches for the operation of filters of spam by the leading providers of internet infrastructure (ISPs) The increasing quantity of unnecessary bulk email (also called spam) has generated a secure need Filters for anti-spam. Then the review compares the strengths and disadvantages of existing methods of machine learning and open research Spam handling problems. As future strategies suggested extreme leaning and strongly opposed schooling that can handle the danger of spam emails effectively.

An Immunological-Based Simulation: A Case Study of Risk Concentration for Mobile Spam Context Assessment

Over the past two decades, there has been a substantial increase in spam messages that caused critical impact loss. Besides the factor of integration of Internet and mobile technology, this issue is also due to the human’s reaction towards spam. This paper presents RiCCA or Risk Concentration for Context Assessment model that performs a risk classification of text spam messages in Short Message Service (SMS) format. The identified risk levels will assist users in anticipating the potential impact of the spam message that they have been receiving. Danger Theory, a prominent theory from Artificial Immune Systems (AIS), inspires the developed model. During the simulation phase, an immunological-based testing lifecycle is applied, with the deployment of the dataset that is shared at UCI Machine Learning Repository and self-collected messages. The performance of the testing revealed a distinctive result, which more than 80% of true positive rate is achieved, employed with two variants algorithm from the Danger Theory; Dendritic Cell Algorithm (DCA) and Deterministic Dendritic Cell Algorithm (dDCA). This simulation demonstrated that the Danger Theory as a feasible model to be applied in measuring the risk of spam. The further articulation on how this immunological-based testing lifecycle is applied in computer simulation and adopting mobile spam as the case study is clarified thoroughly.

인스타그램 기반의 전이학습과 게시글 메타 정보를 활용한 페이스북 스팸 게시글 판별

This study develops a text spam filtering system for Facebook based on two variable categories: keywords learned from Instagram and meta-information of Facebook posts. Since there is no explicit labels for spam/ham posts, we utilize hash tags in Instagram to train classification models. In addition, the filtering accuracy is enhanced by considering meta-information of Facebook posts. To verify the proposed filtering system, we conduct an empirical experiment based on a total of 1,795,067 and 761,861 Facebook and Instagram documents, respectively. Employing random forest as a base classification algorithm, experimental result shows that the proposed filtering system yield 99% and 98% in terms of filtering accuracy and F1-measure, respectively. We expect that the proposed filtering scheme can be applied other web services suffering from massive spam posts but no explicit spam labels are available.

A New Machine Learning based Approach for Text Spam Filtering Technique

A New Machine Learning based Approach for Text Spam Filtering Technique

Clustering-based Spam Image Filtering Considering Fuzziness of the Spam Image

If there are pros, corns are always there. As email becomes a part of individual’s need in our busy life with its benefits, it has negative aspect too by means of email spamming. Nowadays images with embedded text called image spamming have been used by the spammers as effective text spam filtering methods already been introduced. Tracking and stopping spam become challenge in the internet world because of versatility in the spam images. In this paper a novel model AFSIF (Autonomous Fuzzy Spam Image Filter) has been introduced. The basic idea behind AFSIF is, an spam image can combine several basic features of different spam images, so feature fusion weight of the image has been generated, which keeps combined feature of spam images and user preference as well. Here user preference has not been applied separately; it is used to calculate the fusion weight in terms of predefined topics (rule table).

A hybrid spam detection method based on unstructured datasets

The identification of non-genuine or malicious messages poses a variety of challenges due to the continuous changes in the techniques utilised by cyber-criminals. In this article, we propose a hybrid detection method based on a combination of image and text spam recognition techniques. In particular, the former is based on sparse representation-based classification, which focuses on the global and local image features, and a dictionary learning technique to achieve a spam and a ham sub-dictionary. On the other hand, the textual analysis is based on semantic properties of documents to assess the level of maliciousness. More specifically, we are able to distinguish between meta-spam and real spam. Experimental results show the accuracy and potential of our approach.

Image Content-Based “Email Spam Image” Filtering

 Abstract—With the population of Internet around the world, email has become one of the main methods of communication among people. Due to the flood of online information, great amount of the spam emails brings troubles to people. The number of technical approaches of spam filtering are increasing which are mostly based on the text spam filtering technologies. But it is not very effective for test messages imbedded into images which are developing rapidly in recent years. In this paper, we propose an approach of spam image filtering. Our approach combines the characteristics of spam images with the corner point density to detect spam images. The effectiveness of the proposed approach is experimentally evaluated.

Filtering Image Spam: A Survey

Electronic mail being one of the most prevalent methods for exchange of digital messages has been facing the biggest threat that is spam. Since the text spam can be easily detected so a new variant of spamming came into being. The new variant of spam is the image spam that is the trending method of spamming. This paper investigates the several classifiers used in the image spam classification like the Decision Tree and the Support Vector Machine (SVM) and a combination of the both.

A Novel Method of Text Clustering for Chinese Spam Based on Semantic Body

The effect of spam filtering method based on statistics is not good in filtering the new-type spam with synonymous substitution and camouflage. So a new text clustering method based on Semantic Body for filtering Chinese spam is proposed. In this paper, the word sense disambiguation, lexical chain based on HowNet and statistic-based TFIDF are adopted to extract features of mails. The Semantic Body is obtained from the process. The text clustering based on semantic distance is utilized to dealing with Semantic Body. The experimental results under CCERT Chinese-rules.cf show that the proposed approach has a good performance for new type Chinese text spam in filtering.

Spam influence on business and economy: Theoretical and experimental studies for textual anti-spam filtering using mature document processingand naive Bayesian classifier

Spam is unsolicited bulk messages sent indiscriminately. According to Wikipedia and Cisco report, more than 31 trillion spams have been sent in 2009. These spam or “junk mails” can involve various kinds of messages such as commercial advertising, pornography, viruses, doubtful product, get rich quick scheme or quasi legal services. In this paper, a direct attention has been paid to the text spam, and in particular, the process of text spam and the tricks of the spammers have been described in this paper. Moreover, the author described the implementation of the text content analysis and classification, using different document processing techniques (that is, stop words, short words form, regular expression, stemming etc.) and naive Bayesian classifier. In addition to that, the author has depicted the practical work of the document processing and naive Bayesian classifier towards implementing an accurate anti-spam system. Key words: Text spam, stop words, short words form, regular expression, stemming, document processing, naive Bayesian classifier.

Set Cover Feature Selection for Text Categorisation and spam detection

In this paper the performance of the Set Cover (SC) Feature Selection (FS) method for Text Categorisation (TC) and Spam Detection problems is studied. Several variants of the original method are presented either to overcome the drawback of the unbalanced problems which are usually present in TC or to increase the efficiency. The behaviour of the algorithm is tested on several collections. The experiments show these methods provide a great reduction in the dimensionality of the problem either keeping the effectiveness of the classification or causing just a slight decrease.