Monthly Archives: March 2015
I find the efforts by big Web companies to “order the web” fascinating and the implementation of metadata into HTML5 advantageous for the future of utilizing web resources for sophisticated information needs. I found it very intuitive to include this information in html code, since it has been used somewhat similarly for search engine optimization for many years.
The key to the effectiveness of this implementation though lies in the controls placed on descriptive vocabulary. In the case of a traditional collection we typically think of one object type or at least a recognizable format for most, if not all objects. If we can imagine the variety of material presented across the internet though we can quickly see the importance of controlling descriptions through a controlled vocabulary.
This article, why dated somewhat, resonates for several reasons. Most profoundly because this issue is still under scrutiny, though some measures have been enacted since the hearings described in the article. What drew me to this article particularly is the vision expressed by Sir Tim Berners-Lee, a man whom any computer nerd worth his salt reveres for his contributions to computer technology. The “scifi-esque” manner in which he envisions the future is not a far cry from the direction we see technology heading in 2015 with many ventures exploring virtual reality or augmented reality technology. What is striking to me though, through the lens of LIS, is the required sophistication of data management required to maintain accuracy in an always on world. Some of the scenarios in the article seem to be very desirable, but as we have already seen with “smart” appliances such as refrigerators and microwaves these devices are vulnerable to outside attacks. With a need for accurate information that is also stored securely it will be increasingly imperative for information professionals to have a guiding presence in development of storage and retrieval policies and practices.
My DC element is Type and, admittedly, for this particular project it is a straightforward choice both in terms of our choice to keep it and in its usage for this application. DCMI provides a vocabulary for Type which can be reviewed here:
Reviewing this vocabulary it is immediately obvious that for this application we will use the “Image” label for all of our photos. We may find the additional labels “StillImage” or “MovingImage” applicable to some of our material, but ultimately these two labels are to be used under the auspices of the broader label “Image”.
While the distinction in our application is very straightforward, it is easy to understand how this element can be more complicated in a diversified collection of objects. At around 12 labels, the DCMI vocabulary may seem succinct, but if you review the labels and what they are intended to encompass they really cover almost any object conceivable. In this element the effectiveness of the DCMI to allow simple description of objects is very prevalent.
I found the examination of image analysis helpful going into our discussion of what elements to keep for the image collection. I feel like two area examined were already something we have considered – namely “Levels of Analysis” and “Vocabulary” control. The other area of examination “The focus on relationships, processes” struck me as pertinent and unnoticed until I read this article. Quite simply it states that many instances of images derive their “aboutness” meaning not simply from the objects or subjects pictured, but from the relationship the two objects have between one another. In the example given in the article the picture considered is a stork carrying a baby elephant. Separate they have meaning, but the meaning they carry together and then the meaning is only understood culturally. Ins electing which elements to keep and those to dispose of it seem pertinent then that we consider all the information users would need to have for the correct contextual understanding of the image.
When we received our elements I at first considered I had made a folly in choosing Type. Since this collection is all images, what use would my element be? “It says Images in the title”, I thought to myself. In reading the DCMI definition of Type closely though I was struck by the advice, “…best practice to ensure interoperability is to include at least one general type term from the DCMIType vocabulary in addition to the domain specific type term(s)…” It occurred to me I was thinking only locally in my dismissive thoughts on Type and neglecting to consider that these images would be of interest to other institutions which may desire to utilize them within the parameters of their institutions policies. I believe now that utilizing DCMI Type vocabulary – in addition to any local descriptions of Type we may deem relevant and useful – will ensure that the collection can be integrated with outside systems efficiently and effectively. Further I think the Type terms “Image” and “Event” are both applicable to efficient utilization of the image collection. Before using the “Event” term though, further explication of proper usage would be necessary.
I read this article in a recent issue of Forbes magazine. It discusses speaker manufacturer Jawbone and their foray into wearable technology. Their product line includes a wrist device that tracks the user producing information on steps taken and sleep patterns. Analyzing the collected data the company was able to determine it’s accuracy sufficient and started directing guidance to individual users to change their behavior patterns. The article goes on further to discuss the immense amount of resource put into studying behavior in an attempt at creating the most optimal design for the desired user behavior. In the previous post I was looking at the author railed on about the potential thought police/Big Brother effect of the government potentially using collected metadata to attempt to predict behavior, but big companies like Google are already doing this with the data freely given to them.
This post is actually about two articles. I must admit when I first read them I was a little skeptical even though I agree that the measure in Australia is generally not a positive development. The point that is made in the second article though relates the idea that with such a vast hoard of metadata the government will be able to make predictions about behavior patterns and anticipate actions. At this point I am picturing a large warehouse of grey clothed workers at computer terminals assigning tags to phone numbers or website such as terrorist, criminal, et al and a list of people to observe based on their association with that phone number or website. In other words an anticipatory surveillance or suspicion of people based on the relation of their metadata to someone else’s metadata. While I think these articles may be taking an extreme leap somewhat and their tone is definitely editorial I think the concern they raise issue with is valid.
Here we have more news about the metadata political battle ongoing in Australia. I found it interesting that the cost for the program is described as $188 million to $319 million. This struck me because that is a huge gulf between already huge sums of money. Yet, again, no indication of how this data will be managed is given in detail other than to mention that passwords, PINs, et al will not be held by the telecommunication company which is tasked with retention of this data for a period of two years. In response it describes a campaign designed to attempt to flood the attorney general’s email inbox – ironic because this mimics a denial of service attack which is a commonly deployed illicit tactic. Further irony exists I think between the supposed reason for this measure – the prevention of cyber terrorism acts – and the fact that it creates a target for illicit groups in search of the information being retained for criminal reasons.
I am transfixed with controlled vocabularies both their creation and their function. As such this overview was fascinating to me. Even a controlled vocabulary can become out of control, but the principles discussed here represent sound practice for this endeavor. I have made attempts at developing controlled vocabularies in the past and found this the most difficult task. Often I found myself becoming too specific and not, as a previous classmates blog post noted, maintaining a requisite level of ambiguity. I think this would be most difficult to achieve with digital preservation, particularly mass digitization at large scale.
Slistopher posed the following questions recently:
1. What do you define as “tech skills”?
2. Which types of “tech skills” (broadly defined) have you found to be particularly useful in your work environments?
3. What do you see as being the most useful technology skills for LIS professionals, in general?
4. Have you ever taken the time to learn a particular skill, only to find that it wasn’t helpful at all?
5. How important do you think “soft skills” (communication, interpersonal skills, etc.) are for LIS professionals?
6. If you were in the position of hiring a librarian/archivist/other LIS professional, which types of “tech” knowledge/experience would you expect a candidate to possess?
1. ‘Tech skills’ fall into two categories: hardware and software. While some may be more adept at configuring hardware than others and likewise some may feel more competent with a certain program than others, essentially either of these areas of skill are learned and familiarity induced through interaction over time. Many people ask me personally (concerning building computers) “how did you learn all this stuff?” – and the answer is I have been building computers in some capacity for 18 years. Like most things in life, there is no replacement for experience.
2. Software skills are undoubtedly more valued and useful everyday. Understanding hardware is helpful, but not requisite for most everyday simple tasks. Understanding the essentially simple mechanics of programs makes learning foreign programs more expedient.
4. I cannot think of one. Learning any computer or tech skill is beneficial, typically it is the conceptual context of the skill that is the aspect with the most utility.
5. Extremely. Effective communication is often the most decisive factor in success or failure, in my experience.
6. I would look for the skills mentioned in answer #3, at the very bottom of the totem pole is at least adeptness with typical productivity software (MS Office, Adobe, et al).
That’s my two cents!