Viral Marketing Applications Research Articles

Supplementary Influence Maximization Problem in Social Networks

Due to important applications in viral marketing, influence maximization (IM) has become a well-studied problem. It aims at finding a small subset of initial users so that they can deliver information to the largest amount of users through the word-of-mouth effect. The original IM only considers a singleton item. And the majority of extensions ignore the relationships among different items or only consider their competitive interactions. In reality, the diffusion probability of one item will increase when users adopted supplementary products in advance. Motivated by this scenario, we propose a supplementary independent cascade (IC) and discuss the supplementary IM problem. Our problem is NP-hard, and the computation of the objective function is #P-hard. We notice that the diffusion probability will change when considering the impact of its supplementary product. Therefore, the efficient reverse influence sampling (RIS) techniques cannot be applied to our problem directly even though the objective function is submodular. To address this issue, we utilize the sandwich approximation (SA) strategy to obtain a data-dependent approximate solution. Furthermore, we define the supplementary-based reverse reachable (SRR) sets and then propose a heuristic algorithm. Finally, the experimental results on three real datasets support the efficiency and superiority of our methods.

A GENERALIZATION OF THRESHOLD-BASED AND PROBABILITY-BASED MODELS OF INFORMATION DIFFUSION

Diffusion of information through complex networks is of interest in studies such as propagation prediction and influence maximization, both of which have applications in viral marketing and rumor controlling. There are a variety of information diffusion models, all of which simulate the adoption and spread of information over time. However, there is a lack of understanding of whether, despite their conceptual differences, these models represent the same underlying generative structures. For instance, if two different models utilize different conceptual mechanisms, but generate the same results, does the choice of model matter? A classification of diffusion of information models is developed based on the neighbor knowledge of the model infection requirement and the stochasticity of the model. This classification allows for the identification of models that fall into each respective category. The study involves the analysis of the following agent-based models on directed scale-free networks: (1) a linear absolute threshold model (LATM), (2) a linear fractional threshold model (LTFM), (3) the independent cascade model (ICM), (4) Bass-Rand-Rust model (BRRM) (5) a stochastic linear absolute threshold model (SLATM) (6) a stochastic fractional threshold model (SLFTM), and (7) Dodds–Watts model (DWM). Through the execution of simulations and analysis of the experimental results, the distinctive properties of each model are identified. Our analysis reveals that similarity in conceptual design does not imply similarity in behavior concerning speed, final state of nodes and edges, and sensitivity to parameters. Therefore, we highlight the importance of considering the unique behavioral characteristics of each model when selecting a suitable information diffusion model for a particular application.

Computational models for commercial advertisements in social networks

Identifying noteworthy spreaders in a network is essential for understanding the spreading process and controlling the reach of the spread in the network. The nodes that are holding more intrinsic power to extend the reach of the spread are important due to demand for various applications such as viral marketing, controlling rumor spreading or getting a better understanding of spreading of the diseases. As an application of viral marketing, maximization of the reach with a fixed budget is a fundamental requirement in the advertising business. Distributing a fixed number of promotional items for maximizing the viral reach can leverage influencer detection methods. For detecting such “influencer” nodes, there are local metrics such as degree centrality (mostly used as in-degree centrality) or global metrics such as k-shell decomposition or eigenvector centrality. All the methods can rank graphs but they all have limitations and there is still no de-facto method for influencer detection in the domain.In this paper, we propose an extendedk-shell algorithm which better utilizes the k-shell decomposition for identifying viral spreader nodes using the topological features of the network. We use Susceptible–Infected–Recovered model for the simulations of the spreading process in real-life networks and the simulations demonstrates that our approach can reach to up to 35% larger crowds within the same network, with the same number of initial spreaders.

Effective influence estimation in twitter using temporal, profile, structural and interaction characteristics

Influence diffusion is extensively studied in social networks for product or service promotion and viral-marketing applications. This paper proposes two models for social influence estimation, namely Time Decay Features Cascade Model (TDF-C) and Time Decay Features Cascade Threshold Model (TDF-CT). These models overcome three main existing challenges - first, measure the strength of user's influence as an influencer; second, identify the set of users influenced by an influencer; third, estimate the time frame of the influence. TDF-C is an M-TAP based diffusion model, which learns influence probabilities between users using four types of features, namely temporal, interaction, structural, and profile features, and uses Independent Cascade (IC) model for influence estimation. TDF-CT is an extension of the TDF-C model, which uses temporal and interaction features to calculate the diffusion through the Progressive Feedback Estimation (PFE) model in place of IC model. PFE model is a fusion of two diffusion models, i.e., Linear Threshold (LT) and Independent Cascade. TDF-CT handles the limitations of the contemporary diffusion models, i.e., IC and LT. The efficacy of proposed models is evaluated with respect to existing models Independent Cascade (IC), Time Constant Cascade (TC-C), Time Decay Cascade (TD-C), and Time-Depth Decay Cascade (TDD-C). Experimental evaluation over two benchmark datasets namely Darwin and MelCup17 reveal that proposed models are able to make the predictions very close to the real-time in a given time frame. TDF-CT and TDF-C are most suitable for applications requiring high precision and high recall, respectively. Results of spread shape establish the efficacy of models to spread the influence with good coverage of the social network. Results are obtained with improved accuracy by up to 39%.

Deep Reinforcement Learning-Based Approach to Tackle Topic-Aware Influence Maximization

Motivated by the application of viral marketing, the topic-aware influence maximization (TIM) problem has been proposed to identify the most influential users under given topics. In particular, it aims to find k seeds (users) in social network G, such that the seeds can maximize the influence on users under the specific query topics and diffusion model such as independent cascade (IC) or linear threshold (LT).This problem has been proved to be NP-hard, and most of the proposed techniques suffer from the efficiency issue due to the lack of generalization. Even worse, the design of these algorithms requires significant specialized knowledge which is hard to be understood and implemented. To overcome these issues, this paper aims to learn a generalized heuristic framework to solve TIM problems by meta-learning. To this end, we first propose two topic-aware social influence propagation models based on IC and LT model, respectively, which is conducive to better advertising injections. We then encode the feature of each node by a vector and introduce a model, called deep influence evaluation model , to evaluate the user influence under different circumstances. Based on this model, we can construct the solution according to the influence evaluations efficiently, rather than spending a high cost to compute the exact influence by considering the complex graph structure. We conducted experiments on generated graph instances and real-world social networks. The results show the superiority in performance and comparable quality of our framework.

Budgeted Profit Maximization Under the Multiple Products Independent Cascade Model

Influence maximization problem asks for a small subset of nodes in a social network that could maximize the spread of influence, which finds important applications in viral marketing. Most existing works focus on maximizing the influence of a single product or products that are in pure competition. However, a company usually produces different products to meet the needs of different people in reality. In this paper, we focus on the propagation of multiple products and propose multiple products independent cascade (MPIC) model, which allows each user to adopt multiple products. Aiming to maximize the overall profit across all products, we study the budgeted profit maximization (BPM) problem. To give a high-quality solution for BPM problem under the MPIC model, we present a modified greedy algorithm and derive the performance guarantee in doing so. Furthermore, we show that the two-phase profit maximization algorithm can not only handle the large-scale networks but also give the same approximation ratio as the modified greedy. In addition, we propose the cost performance update heuristics algorithm that has the results close to the above algorithms, and the running time is less than one ten-thousandth of the greedy. Our experiments on three real datasets verify the correctness and effectiveness of our methods, as well as the advantage of our methods against the traditional methods.

CBIM-RSRW: An Community-Based Method for Influence Maximization in Social Network

Influence maximization is an important problem, which seeks a small set of key users who spread the influence widely into the network. It finds applications in viral marketing, epidemic control, and assessing cascading failures within complex systems. The current studies treat nodes in social network with equal weights, and the influence possibility mainly decide by node degree. In this paper, we study the influence maximization problem in social networks and we improve the independent cascade model to realize the goal of different weights for different users, and the differentiation of influence probability. Meanwhile, We take advantage of the community structure to speed up the algorithm. Then, we propose a method called the reverse reachable index method based on random walk (RSRW) to select potential high-impact nodes from those communities. The experimental result on four actual data set shows that these improvements can greatly reduce the calculation time while ensuring the accuracy of the results.

ПІДВИЩЕННЯ ЕФЕКТИВНОСТІ ВІРУСНОГО МАРКЕТИНГУ ЗА ДОПОМОГОЮ ВІЗУАЛЬНОГО КОНТЕНТУ

An analysis of domestic and foreign experience of the use of content-marketing tools, and its creative nature - viral marketing has been analyzed. The most effective tools of content marketing are systematized and its main directions of application in the practical activity of modern Ukrainian business are determined. The disadvantages and advantages of viral marketing, in particular buzz marketing, are considered. It has been determined that with the help of viral marketing, advertising information is distributed in the form of «Sarafan Radio», and the recipients of the advertising appeal themselves are the carriers. The main components of the virus content are presented, their characteristics and the scheme are given - technology of information dissemination. The specificity and effectiveness of the application of viral marketing has been determined, thanks to the rapid chain reaction of the distribution of advertising appeal. The necessity of introducing visual content into buzz marketing as a high-efficiency means of marketing communications is substantiated.

Identification of influential spreaders in complex networks using HybridRank algorithm

Identifying the influential spreaders in complex networks is crucial to understand who is responsible for the spreading processes and the influence maximization through networks. Targeting these influential spreaders is significant for designing strategies for accelerating the propagation of information that is useful for various applications, such as viral marketing applications or blocking the diffusion of annoying information (spreading of viruses, rumors, online negative behaviors, and cyberbullying). Existing methods such as local centrality measures like degree centrality are less effective, and global measures like closeness and betweenness centrality could better identify influential spreaders but they have some limitations. In this paper, we propose the HybridRank algorithm using a new hybrid centrality measure for detecting a set of influential spreaders using the topological features of the network. We use the SIR spreading model for simulating the spreading processes in networks to evaluate the performance of our algorithm. Empirical experiments are conducted on real and artificial networks, and the results show that the spreaders identified by our approach are more influential than several benchmarks.

How to choose friends strategically

Alice wants to join a new social network, and influence its members to adopt a new product or idea. Each person v in the network has a certain threshold t(v) for activation, i.e. adoption of the product or idea. If v has at least t(v) activated neighbors, then v will also become activated. If Alice wants to make k new friends in the network, and thereby activate the most number of people, how should she choose these friends? We study the problem of choosing the k people in the network to befriend, who will in turn activate the maximum number of people. This Maximum Influence with Links Problem has applications in viral marketing and the study of epidemics. We show that the solution can be quite different from the related and widely studied influence maximization problem where the objective is to choose a seed or target set with maximum influence. We prove that the Maximum Influence with Links problem is NP-complete even for bipartite graphs in which all nodes have threshold 1 or 2. In contrast, we give polynomial time algorithms that find optimal solutions for the problem for trees, paths, cycles, and cliques.

Whom to befriend to influence people

Alice wants to join a new social network, and influence its members to adopt a new product or idea. Each person v in the network has a certain threshold t(v) for activation, i.e. adoption of the product or idea. If v has at least t(v) activated neighbors, then v will also become activated. If Alice wants to activate the entire social network, whom should she befriend? More generally, we study the problem of finding the minimum number of links that a set of external influencers should form to people in the network, in order to activate the entire social network. This Minimum Links Problem has applications in viral marketing and the study of epidemics. Its solution can be quite different from the related and widely studied Target Set Selection problem. We prove that the Minimum Links problem cannot be approximated to within a ratio of O(2log1−ϵ⁡n), for any fixed ϵ>0, unless NP⊆DTIME(npolylog(n)), where n is the number of nodes in the network. On the positive side, we give linear time algorithms to solve the problem for trees, cycles, and cliques, for any given set of external influencers, and give precise bounds on the number of links needed. For general graphs, we design a polynomial time algorithm to compute size-efficient link sets that can activate the entire graph.

Revisiting the stop-and-stare algorithms for influence maximization

Influence maximization is a combinatorial optimization problem that finds important applications in viral marketing, feed recommendation, etc. Recent research has led to a number of scalable approximation algorithms for influence maximization, such as TIM + and IMM , and more recently, SSA and D-SSA . The goal of this paper is to conduct a rigorous theoretical and experimental analysis of SSA and D-SSA and compare them against the preceding algorithms. In doing so, we uncover inaccuracies in previously reported technical results on the accuracy and efficiency of SSA and D-SSA , which we set right. We also attempt to reproduce the original experiments on SSA and D-SSA , based on which we provide interesting empirical insights. Our evaluation confirms some results reported from the original experiments, but it also reveals anomalies in some other results and sheds light on the behavior of SSA and D-SSA in some important settings not considered previously. We also report on the performance of SSA-Fix , our modification to SSA in order to restore the approximation guarantee that was claimed for but not enjoyed by SSA . Overall, our study suggests that there exist opportunities for further scaling up influence maximization with approximation guarantees.

Identification of influential spreaders in online social networks using interaction weighted K-core decomposition method

Abstract Online social networks (OSNs) have become a vital part of everyday living. OSNs provide researchers and scientists with unique prospects to comprehend individuals on a scale and to analyze human behavioral patterns. Influential spreaders identification is an important subject in understanding the dynamics of information diffusion in OSNs. Targeting these influential spreaders is significant in planning the techniques for accelerating the propagation of information that is useful for various applications, such as viral marketing applications or blocking the diffusion of annoying information (spreading of viruses, rumors, online negative behaviors, and cyberbullying). Existing K-core decomposition methods consider links equally when calculating the influential spreaders for unweighted networks. Alternatively, the proposed link weights are based only on the degree of nodes. Thus, if a node is linked to high-degree nodes, then this node will receive high weight and is treated as an important node. Conversely, the degree of nodes in OSN context does not always provide accurate influence of users. In the present study, we improve the K-core method for OSNs by proposing a novel link-weighting method based on the interaction among users. The proposed method is based on the observation that the interaction of users is a significant factor in quantifying the spreading capability of user in OSNs. The tracking of diffusion links in the real spreading dynamics of information verifies the effectiveness of our proposed method for identifying influential spreaders in OSNs as compared with degree centrality, PageRank, and original K-core.

The Influence of Customers Communication Behaviour in the Implementation of 21cineplex Viral Marketing Using the Social Networking Site Facebook

Application of viral marketing through social networking sites will greatly help companies in saving high marketing costs since companies do not have to give financial incentive to internet users or customers who voluntarily forward the information about the companies. Good marketing communication strategy will increase the company's branding and enable good interaction between customers and companies. This research to look if the customer communication behavior has a relationship and a significant impact on forwarding online content and whether the curiosity variable has a relationship and a significant impact on the variable of consumption of online content in the application of Viral Marketing. The methodology used is interpersonal communication model adopted from FIRO theory, which consists of the following variables: inclusion-need to belong, inclusion-individuation, affection-altruism, control-personal growth, and consumption of online content, towardthe online content forwarding variable; and also the curiosity variable towards the consumption of online content variable, where the hypothesis was tested by the correlation and regression analysis. The test results based on correlation analysis show that all H1hypotheses (have relation) are received; and based on regression analysis, all H1hypotheses (effect) are received, although the value of r and the effect is less significant (small).

Viral Marketing in the Food Industry: Canada vs. Serbia

The authors explore the application of viral marketing in the food industry in Canada (as a representative of a high-income country) vs. Serbia (a middle-income country). A theoretical background on the scope of viral marketing is given at the beginning. The initial secondary research pointed out that there was a great number of viral marketing campaigns in the food industry, in both high- and middle-income countries. The primary research showed a high degree of use of viral marketing in the food industry in general, as well as a number of similarities between the two countries. Positive experiences with viral marketing in the food industry were far more common than negative experiences. Most marketing professionals, in both countries, believed that a viral marketing program would benefit their company, and they planned to run a viral marketing campaign again in the future.

“With a little help from new friends”: Boosting information cascades in social networks based on link injection

We investigate information cascades in the context of viral marketing applications. Recent research has identified that communities in social networks may hinder cascades. To overcome this problem, we propose a novel method for injecting social links in a social network, aiming at boosting the spread of information cascades. Unlike the proposed approach, existing link prediction methods do not consider the optimization of information cascades as an explicit objective. In our proposed method, the injected links are being predicted in a collaborative-filtering fashion, based on factorizing the adjacency matrix that represents the structure of the social network. Our method controls the number of injected links to avoid an “aggressive” injection scheme that may compromise the experience of users. We evaluate the performance of the proposed method by examining real data sets from social networks and several additional factors. Our results indicate that the proposed scheme can boost information cascades in social networks and can operate as a “people recommendations” strategy complementary to currently applied methods that are based on the number of common neighbors (e.g., “friend of friend”) or on the similarity of user profiles.

Large-scale evaluation framework for local influence theories in Twitter

Influence theories constitute formal models that identify those individuals that are able to affect and guide their peers through their activity. There is a large body of work on developing such theories, as they have important applications in viral marketing, recommendations, as well as information retrieval. Influence theories are typically evaluated through a manual process that cannot scale to data voluminous enough to draw safe, representative conclusions. To overcome this issue, we introduce in this paper a formalized framework for large-scale, automatic evaluation of topic-specific influence theories that are specialized in Twitter. Basically, it consists of five conjunctive conditions that are indicative of real influence exertion: the first three determine which influence theories are compatible with our framework, while the other two estimate their relative effectiveness. At the core of these two conditions lies a novel metric that assesses the aggregate sentiment of a group of users and allows for estimating how close the behavior of influencers is to that of the entire community. We put our framework into practice using a large-scale test-bed with real data from 75 Twitter communities. In order to select the theories that can be employed in our analysis, we introduce a generic, two-dimensional taxonomy that elucidates their functionality. With its help, we ended up with five established topic-specific theories that are applicable to our settings. The outcomes of our analysis reveal significant differences in their performance. To explain them, we introduce a novel methodology for delving into the internal dynamics of the groups of influencers they define. We use it to analyze the implications of the selected theories and, based on the resulting evidence, we propose a novel partition of influence theories in three major categories with divergent performance.

Maximizing the Cumulative Influence through a Social Network when Repeat Activation Exists

We study the problem of employing social networks for propagate influence when repeated activations are involved. While influence maximization has been extensively studied as the fundamental solution, it neglects the reality that a user may purchase a product/service repeatedly, incurring cumulative sales of the product/service. In this paper, we explore a new problem of cumulative influence maximization that brings the influence maximization a step closer to real-world viral marketing applications. In our problem setting, repeated activation exists and we aim to find a set of initial users, through which the maximal cumulative influence can be stimulated in a given time period. To describe the repeated activation behavior, we adopt the voter model to reflect the variation of activations over time. Under the voter model, we formulate the maximization problem and present an effective algorithm. We test and compare our method with heuristic algorithms on real-world data sets. Experimental results demonstrate the utility of the proposed method.

THE ATTITUDES TOWARD APPLICATION OF VIRAL MARKETING IN THE FOOD INDUSTRY IN SERBIA

This paper examines the attitudes toward application of viral marketing in the food industry in Serbia. The research consisted of both an extensive theory review and empirical research, including case studies, surveys and in-depth interviews. Viral marketing has been defined as any marketing program (online or offline) that is designed to achieve an exponential growth by spreading marketing effects from customer to customer. The paper’s hypothesis, stating that marketing managers in Serbia have positive attitudes toward the usefulness of viral marketing in the food industry, has been supported by the results of both primary and secondary research. Finally, a number of limitations and risks associated with the viral marketing strategy have been presented, followed by a brief discussion of possible ways to overcome the challenges.

A note on maximizing the spread of influence in social networks

We consider the spread maximization problem that was defined by Domingos and Richardson (2001, 2002) [7,22]. In this problem, we are given a social network represented as a graph and are required to find the set of the most “influential” individuals that by introducing them with a new technology, we maximize the expected number of individuals in the network, later in time, that adopt the new technology. This problem has applications in viral marketing, where a company may wish to spread the rumor of a new product via the most influential individuals in popular social networks such as Myspace and Blogsphere. The spread maximization problem was recently studied in several models of social networks (Kempe et al. (2003, 2005) [14,15], Mossel and Roch (2007) [20]). In this short paper we study this problem in the context of the well studied probabilistic voter model. We provide very simple and efficient algorithms for solving this problem. An interesting special case of our result is that the most natural heuristic solution, which picks the nodes in the network with the highest degree, is indeed the optimal solution.