Discussion Thread on IoT

In response to the question posed about whether data from the Internet of Things should be meaningfully linked, a recent article (https://thenewstack.io/w3c-linked-data-help-standardize-internet-things/) outlines how this is currently a central priority for the W3C.  The W3C considers Linked Data key to standardizing the IoT. To address the importance of LOD to the development of the IoT, W3C has created a "web of things" working group "to bring together all the competing interests, protocols and standards to create recommendations for...the Internet of Things (IoT)." At the heart of the W3C's work with the IoT is to enable semantic interoperability which allows for "consistency and fluency...across standards and platforms." Drawing from this week's reading, "Semantics for the Internet of Things: early progress and back to the future"(Barnaghi et al., 2012), it is important to keep in mind the "diversity, heterogeneity and spatiotemporal dependency of IoT data and resources" and the obstacles posed by the current IoT which is largely disconnected and inconsistent at semantic levels. For the W3C, creating standardized semantic frameworks using LOD is "essential to describe and represent the data and to make it seamlessly accessible and process-able across different domains."

I also came across the same article that Maria mentioned. Data that IoT generates is mostly real-time, and need to be continuously updated to reflect the physical world. The major issues with IoT technology is fragmentation. Data across different platforms, and systems can not communicate with each other; therefore, they need to be linked meaningfully at semantic level. Moreover, "Things" in the IoT can be anything, any objects and devices. Thus, ontologies, semantic annotation, Linked Data and semantic Web services are considered as principal solutions. "Providing unambiguous data descriptions in a way that can be processed and interpreted by machines and software agents is a key enabler of automated information communications and interactions in IoT" (Barnaghi et al, 2012).

Barnaghi, P., Wang, W., Henson, C., & Taylor, K. (2012). Semantics for the Internet of Things: early progress and back to the future. International Journal on Semantic Web and Information Systems (IJSWIS), 8(1), 1-21.

Excellent point Nguyen. I came across several articles when browsing for a selection for the mini assessment that made the point that sensor devices and their data are often closed systems. This leads to the inevitable conclusion of fragmentation... Semantic Web technologies seem to be a pretty promising solution to this, especially since they've proven their value in other areas. Here's an article I found pretty interesting on the subject. It discusses how to use semantic web technologies to integrate sensor data:

http://ai2-s2-pdfs.s3.amazonaws.com/fd68/69e660b0ced599c62d2f58128c64e59e5967.pdf

Nicholas your question is main issue as I see it facing IoT: how to standardize the applications, systems, and data that IoT "Things" use and create. How do we get Big Company 1 and Big Company 2 to use the same operating and data systems so that when the repair person tries to fix the "Thing" he or she can access the "Thing's" data without a computer science degree. Somehow, W3C is going to have to find a way to motivate a few of the big companies to adopt its systems and hope that it turns into an "industry" standard that all the manufacturers of IoT "Things" use. 

Also, I question whether it will make things better to have all "Things" connected to the Internet. Will it really help the consumer or is it something that is not necessary except to those seeking to profit off it. For example, I bought a new washing machine 5-7 years ago. No problems. Do I need it connected to the Internet? Do I want it connected to the Internet? Short answer, no. On the other hand, self driving cars must be connected. 

I found this article on Matching over linked data streams in the Internet of things which helped me have a better understanding of how quality standards of LOD can be more meaningful, when in the IoT the main data connected is things, whereas the in the Internet of Computers, the main data producers and consumers are humans.

The question is do we want a high-quality meaningful information linked in such a manner that we don’t get false-negative matches?

With the Internet of Things’ ontologies such as the smart product ontology for a shared conceptualisation of product requirements, smart city scenarios will be real world. The technology available now for Smart cars with forward collision warning, lane changing warning and other vehicle management functions linked to the manufacturer, self-driving vehicles an increasing reality!

Qin, Yongrui, Quan Z. Sheng, and Edward Curry. "Matching over linked data streams in the internet of things." IEEE Internet Computing 19, no. 3 (2015): 21-27. 

Yes, I agree with your point. Where devices or the "Things" in IoT involve shared understanding, there's definitely a need for standards and quality supported by those standards. After all, driving fundamentally involves standards (not only rules of the road, but also rules of what a car should be with its emissions, its handling, and its interaction with its environment).

As the IoT continues to generate massive volumes of connected data over the next few years, it is important to consider how the IoT is increasingly being relied on as evidence in legal situations.  For instance, in April of this year, the data produced by a fitbit was used in a trial as evidence to convict a murderer. See NYT article, "In Connecticut Murder Case, a Fitbit Is a Silent Witness".

As the legal community increasingly depends on the IoT for sound digital forensic evidence, the stakes for guaranteeing high quality linked data increase. In "Semantics for the Internet of Things: early progress and back to the future". Barnaghi et al. descrbe how the IoT data provided by sensory devices varies in quality, measurements can change over time, and there can be faults and errors in the devices and their settings. Inaccurate and varying qualities of IoT data pose huge problems for the reliability of IoT data as legal evidence. By using LOD to provide reliable semantic descriptions of quality related attributes of IoT data, we can ensure that the IoT data used in legal situations is reliable and accurate and can be depended on as legal evidence.

The massive accumulation of electronic data, software, devices, and internet platforms, comprising multiple forms and speaking many computer languages, has led to a general lack of standards for linked open data. I think this creates a stronger requirement for LOD creators and publishers to provide meaningful links, because the description of objects within the Internet of Things can no longer be taken for granted or assumed to be the same by everyone. The IoT has made our job a bit more complicated because now we have to integrate various technologies and ontologies into our online databases and collections. 

Wojcik (2016) discusses how libraries can embrace the Internet of Things in order to improve their services and public reputation, like hosting mobile apps to keep users informed of library hours and computer availability, in a manner similar to commercial businesses. But it seems to me that this whole integration process of IoT is too broadly defined (like trying to define the “marketplace”) and that organizations often find themselves stumbling about trying to come up with IoT solutions. Banafa (2016) points out the dichotomy of the IoT becoming critical to our everyday lives while creating security and privacy risks. So the IoT perhaps is something in which we must now navigate or exploit for good quality data, descriptions, services, user connections, etc.

I think the viag.org site for name search is comprehensive, robust, descriptive, and relatively accurate. It provides links for accessing related persons or corporators, URIs, MARC records, and wiki pages for corroborating information. I cannot help but feel that sometimes I would still need to find information elsewhere, in case the links are not accurate or complete.