Art Images Online: Leveraging Social Tagging and Language for Browsing
This small article offers a solution to the problems inherent in tagging as was brought up in the IMLS Leadership Project "T3: Text, Tagging and Trust to Improve Image Access for Museums and Libraries". Namely, the great difference in specificity of the tagging. It offers "computational linguistic processing of user-contributed tags and associated text" and "leveraging multilingual tags" sections.
In the first section it offers their results for the computational linguistic tools. These tools sit at the back end of a website similar to www.steve.museum, and the front end gathers tags. Their tools were able to achieve 79% accuracy in lemmatizing (finding the base of words). They also looked at Parts of Speech Ambiguity, and found that such terms often occur in the same tag cloud, which can be used to "determine cohesion of bigrams". They also determine that most text about images is about the work of art, rather than the context (ie the painter).
The article also discussed that multilingual tags. The most popular terms are often exact translations of the most popular terms in the other language (45% of the time) or the second most popular term in the other language (36.36% of the time). The most common tagging elements where in the following order: "general person or thing", "visual elements" and "non-subject matter" (painter names, etc.)
Implications:
For my dissertation, the tool that lemmitizes tags at such a high rate of accuracy is clearly important, and could definitely help to overcome the perception that tagging offers inconsistent results. The results showing that the most common results of multilingual tagging are often direct translations is interesting in that it offers an argument that tagging leads to standardisation. I will have to take a look at the references of this article given how most of use was found in such a short article.
Enacting engagement online: framing social media use for the museum
This article discusses how museums might want to re-engage with their audiences. In particular it deals with the problem that although museums might want to do this, their methods have traditionally left the public feeling that the museum was trying to take their traditional authority on interpretation into a new sphere. It does this by frame analysis, with the frames: The Marketing Frame (advertising the museum's activities, and promoting the brand); The Inclusivity Frame (providing inclusion for the public into the activities of the museum); The Collaborative Frame (the frame where the museum allows for the audience to co-produce narratives).
One of the problems the article found is that museums often have a "risk averse culture", and this means that their efforts state that they intend to empower, interact with and democratise their processes, but rarely actually do so. Often museums simply launch social media initiatives, without understanding the technical resources needed in order to garner the benefit they hope to achieve (ie, that simply joining Facebook, and having an audience doesn't mean that the posts will automatically achieve any of the museum's goals without planning and staff who know how to engage with the audience in that platform).
The article proposes those three frames as a beginning, but also mentions that there are other potential frames of analysis.
Implications:
The main implication is that museums may state that they are interested in engaging with the general public concerning their collections, but haven't thus far actually done so. They often seem to be a bit preachy, and without developing a good framework for addressing increased public engagement, they run the risk of putting effort into something in a way that doesn't gain them the intended benefit. Perhaps of the greatest implication is the fact that this article was published by someone at my university. So, it might be interesting to see if I can contact this person in order to see if she has further insights or suggestions.
1. What type of cataloguing system does your organisation use for your cultural collection?
a. AACR2
b. CCO
c. Library of Congress Thesaurus for Graphic Materials
d. Hybrid
e. In-house standard
f. other
2. What type of metadata system does your organisation use for your cultural collection?
a. VRA Core
b. CDWA
c. Dublin Core
d. Hybrid
e. In-house standard
f. other
3. How receptive is your organisation to new developments in cataloguing for cultural objects?
5 point likert scale
4. How receptive is your organisation to new developments in metadata for cultural objects?
5 point likert scale
5. Does your organisation use crowdsourcing that you are aware of?
Yes
No
6. How receptive is your organisation to crowdsourcing that you are aware of?
5 point likert scale
7. Does your organisation have legal requirements influencing cataloguing or metadata of cultural objects that you are aware of?
Yes
No
8. Does your organisation have legal requirements influencing use of crowdsourcing that you are aware of?
Yes
No
No comments:
Post a Comment