Machine Learning Based Trust Computational Model for IoT Services

Abstract

The Internet of Things has facilitated access to a large volume of sensitive information on each participating object in an ecosystem. This imposes many threats ranging from the risks of data management to the potential discrimination enabled by data analytics over delicate information such as locations, interests, and activities. To address these issues, the concept of trust is introduced as an important role in supporting both humans and services to overcome the perception of uncertainty and risks before making any decisions. However, establishing trust in a cyber world is a challenging task due to the volume of diversified influential factors from cyber-physical-systems. Hence, it is essential to have an intelligent trust computation model that is capable of generating accurate and intuitive trust values for prospective actors. Therefore, in this paper, a quantifiable trust assessment model is proposed. Built on this model, individual trust attributes are then calculated numerically. Moreover, a novel algorithm based on machine learning principles is devised to classify the extracted trust features and combine them to produce a final trust value to be used for decision making. Finally, our model's effectiveness is verified through a simulation. The results show that our method has advantages over other aggregation methods.

Trust Management in Internet of Things Applications

“…At that point a multi-class characterization procedure like support vector machine (SVM) is utilized to prepare the ML model so as to recognize the best edge level that isolates dependable collaborations from others. [2] Our primary target is to differentiate malicious interactions from trustworthy interactions with with most extreme limit division and least exceptions as opposed to classification itself.…”

Data centric trust evaluation and prediction framework for IOT

“…However, calculating these weighting factors are computationally costly and not practical due to infinite number of possibilities. Hence, we suggest to apply machine learning (ML) techniques to combine all TAs, which we have discussed in our previous work [7].…”

“…where  and  are weighting factors such that + =1 and > 0. The ML method discussed in [7] is preferable for TA combination in this case as well. where represents the reputation towards data source B by its previous users n. A mechanism that computes reputation based on PageRank algorithm is presented in our previous research [2].…”

“…To this end, firstly we define a set of dynamic factors, which essentially describe the DQ attributes and also metrics which define the knowledge, experience and reputation as in our previous work [2] and [1] to get the best of traditional means of trust computation. Then, we combine these attributes built on REK (Reputation, Experience and Knowledge) model described in [4] and [7]. After that, a technique which assesses the data centric trust for every user who is new to the system and who needs to access the data streams, is investigated based on the concepts of recommendation systems (RS).…”

“…Among them, a trust relationship model described in this work elaborates some important factors when it comes to trust based decision making that is a vital part of standardization process. On the other hand, trust evaluation via ensemble methods, which is by combining numerical, machine learning and recommendation algorithms, provides robust perception about trust compared to traditional one dimensional trust calculation techniques [5][6][7]. Additionally, we propose and encourage to use publish-subscribe architecture in the process of data management due to its distributed and autonomous nature.…”