channelAustin feedback on metadata
channelAustin has established the 15 following theme timeslots
Adult / Explicit
Arts / Culture / Entertainment
Community Service
Education
Environment
Ethnic
Experimental
Health
International Issues
Lifestyle
Music
News / Public Affairs / Government
Spiritual / Faith-based
Sports
Youth
Pluse we have some feedback so far on the metadata schema
OK, here are some points and observations.
1) TOPIC Categories
I see 28 topics. IF a goal is to consolidate I can see combining New Age, Spirituality, and Religion. I don't think that the term "New Age" resonates much here in Austin. It seems specific to certain regions of the country -- or more prevalent in certain regions anyway, and maybe out of fashion.
Also, I see Community Affairs and Public Affairs as having a lot of overlap. I'm wondering what would be considered a Public Affairs program that is also NOT a Community Affairs program, and vice versa.
I see Women's Issues is a Topic, but also Women is a Target Audience. Whereas I see Target Audiences for a variety of ethnic/race/region types (Asian/Asianamerican, etc), but not a corresponding Topic. Just an observation, not a critique.
I don't see Dance as a topic. Maybe that would be under Arts.
Lifestyle could be a topic
2) TARGET Audience
I definitely get Adult and Children as target audiences. However, am I to infer that "Adult" implies mature content, i.e. MA, or does it simply mean content that adults might find fascinating, yet kids would be bored to tears over?
I'm not so sure that I or anyone would make a program that is targetted specifically to European/Euro-Am (and is it a typo, or intentionally left as Euro-Am, rather than European-American?)
Regarding Latin/Latin-American, here in Texas among Tejanos (people of mixed Spanish - Indigenous descent whose families go back a few generations) they would not necessarily self-identify or target Latin/Latin-American
"""The majority of Tejanos today are white Hispanics; over 63% of all Hispanics in Texas, in the 2007 ACS, consist mostly of Mexican Americans whose ancestors arrived in Texas prior to and during the Mexican Revolution.""" http://en.wikipedia.org/wiki/Tejano
And here the term Hispanic is widely used.
We have a producer with the show title "Hispanic Today" and another called "Prime Time Tejano"
Obviously if Latin/Latin-American is the only choice on the list, they would pick that as a target audience. But it might not be the best preference.
I apologize if you've already grappled with this. And I don't have a solution, because if you start expanding it to include Hispanic, Mexican-American, etc, then you really grow the list of audiences to be too many -- probably.
Also wonder why there is South Asian as a stand alone category, when there is Asian/Asian-American.
3) STYLES
"Lecture" could be considered a style. Other than that, I don't see anything to add there.
4) MAPPING channelAustin Time slots
We have developed 13 theme timeslots here. And when I attempt to map potential topics, target audiences, and styles, I'm noticing some patterns.
A few examples of our theme timeslots are Arts / Culture / Entertainment, Education, International Issues, Lifestyle . . . Well for these the Target Audience would or could be all on the list in the Target Audience category. For nearly all of the theme timeslots they make sense to map to all audiences.
I suppose that means that we haven't developed a theme timeslot with a narrow base. Which in fact makes me wonder about the utility of having the target audience as part of the metadata.
Why do we need a target audience info? Are we assuming that some people are going to search for content based on the target audience, rather than based on the topic?
Similarly, when mapping Style to timeslot themes, I'm also seeing that many of our timeslot themes could or would be possible in nearly all of the Style categories.
For example for our Arts / Culture / Entertainment theme timeslot, I didn't map it to Cooking, Fix it / How to, or News / Magazine . . . but I can even see how it could map there.
I guess I'm not clear on what we are doing here with this. . . .
5) CONCLUSION
The Topics make the most sense to me, I'm still not clear on Target Audience and Style.
If the theme timeslot chosen determines the list of metadata choices available, why not just have nearly all available.
Clarification on this might help me understand better.
- Stefan
Target Audience and Style
Thanks for the feedback Stefan.
The thought with Target audience was if we had a Theme Recommendation Engine we may need additional information about the show in order to determine which theme block a show best belongs in. For example, you have a show in Austin that is a "entertainment" show but it focuses on Spanish culture and humor and is in a theme block called "Hispanic". In Denver, we'd like to air that show too but we have a theme block called "Cultural Perspectives" where we'd like it to air. However, if it just has the topic "entertainment" it wouldn't go into that block at our station. Another example could be a health show, often they have very specific target audiences: women, GLBTQ, African American... Right now we have GLBTQ and African American as theme blocks and would need content like that to be populated in those specific blocks but looking at topic alone would probably not get content in there.
Similarly, style is supposed to aid in recommending the correct theme for the producer and for any stations sharing content. If someone marked a show with music as a topic but the style of the show wasn't a performance but a talk show it might not end up in the right theme block.
It really depends on what the theme blocks are at any given station and how you think content should fit in your blocks. We are not 100% agreed here either and this is a good discussion that I hope all the stations will weigh in on.
Ann
Gaining clarity on theme timeslot metadata
From off-GDO iChat w/ Kevin, I have a much better understanding now of what is attempting to be achieved with respect to the theme timeslots and the creation of these metadata filters.
Correct me if I'm wrong, but I think the concept is this:
1) Each station will develop their own theme timeslots based on experience and / or anticipated use
2) Each station will assign theme timeslots to an array of metadata choices based on a)Topic b)Target Audience and c) Style
3) When a user submits a program via the Create Show web form, he or she will choose a theme timeslot, and in addition select from the array of metadata choices that best characterize the program
4) It is then this metadata that will accompany the program if or when it is shared across the network to other stations
5) Another station receiving that program will likely not have the same theme timeslot, however it is the "Theme Recommendation Engine" that will examine the existing program metadata and use that to recommend the best match for the receiving stations theme timeslot.
Maybe the process is more nuanced that this, but I think I'm basically correct.
Given that . . . I have a couple of comments.
First of all, I'm noticing that the theme timeslots that we've developed at channelAustin are so broad that in the metadata category of Target Audience and Style it seems appropriate to select nearly all the metadata options on that list.
And therefore it seems that this "Theme Recommendation Engine" would work best when there are a greater number of more specialized or niche theme timeslots.
So, for example, if there were to be a theme timeslot called Green Kids Documentaries then there would be a very narrow set of metadata: Topic = environment (plus others) Target Audience = children, or general, Style = documentaries.
On the flip side, for a theme timeslot called Environment -- which we created here -- the Topic would be environment, and perhaps others, but the Target Audience for a theme as broad as Environment could basically include everyone of the Target Audience categories, and the Style of Environment themes could be nearly any one of styles listed.
And now for a random observation.
On DOM's theme timeslots, I noticed that "Muslims" was mapped to African American theme timeslot, but "Christians" was not. And on the Latino theme timeslot "Christians" was not, but only the broader metadata category of "Religion"
Does this mean that if a producer in Denver submitted a program for the African American theme timeslot about a Baptist Church service that he or she would indicate a metadata category of Religion, whereas if it were a program about a mosque, they could indicate both Religion and Muslims? And a program for the Latino theme timeslot cannot specify Christian, but only Religion?
Maybe I am still confused.
Are all the metadata categories exposed to the user at the time of submitting a program? And are we mapping the metadata to theme timeslots only for the purpose of matching appropriate theme timeslots at another station?
Portland Community Media Timeslot Theme Feedback
Portland Community Media gives our producers the option to choose between 7 broad themes.
These are:
1) Arts & Entertainment
2) Community Interest / Public Affairs/ Live
3) Education
4) Ethnic
5) Health
6) Inspirational
7) Sports
These themes are tied to daily time slots on our programming calendar. If more themes were added, the scheduling would not diversify further. They would just be tied together based on similar themes within those time slot events, which I believe
would not benefit programming and may end up confusing our producers.
This "less is more" approach has been working well in Portland and it allows the flexibility of programs to be scheduled in one or more theme. It also simplifies the selection process for the producer because the themes, although clear-cut, do allow for some cross-over ("ethnic - inspirational"). I could see this potentially causing issues with a Theme Recommendation Engine if you are a station with very specific theme blocks.
The categories do get more specific once programs are ingested into the Princeton Server. These categories are selected by playback staff and were mainly created to assist them in creating a daily schedule or finding filler.
These categories are:
Arts
Childrens
Culture
Education
Entertainment
Ethnic
Filler
Health
Information
Municipal
Other
Potentially Objectionable
Public Affairs
Seniors
Spiritual
Sports
In this case choosing multiple categories helps create more specific information about the program (Filler - Potentially Objectionable).
We also have a Princeton Server dedicated to our government channel and POP 29, which is a channel filled with programs either created by our multi-media services department or outside programming (LinkTV, pegmedia.org, documentaries, creative commons works, TED talks, etc).
The themes on this Princeton Server are designated as follows:
First Tier:
POP 29 or Municipal/ Government
Second Tier:
Series or Individual
Categories:
Art & Culture
Comedy & Drama
Documentaries
Feature
Filler
Friday Night Movie
Home & Garden
Inspirational
Monday Night Documentary
PSA
Public Affairs
Saturday Morning Cartoons
Saturday Night Movie
Sports
Youth & Education.
Some are broad themes and some are tied to recurring weekly programs that are scheduled for broadcast only one day a week.
It does appear that there needs to be a fine balance struck between too few categories, which do not provide enough scheduling information (especially to stations that have already established very specific time slot themes) and the possibility of sharing a program that has detailed theme information tied to it and must then be deciphered by the scheduler to fit into their station's broader theme structure.
In this situation I agree with Stefan, that some of these categories can be combined and simplified. It is also hard for me to imagine that all participating stations will either (A) have the same themes or (B) have producers comb and select their target audience from such a long list. Producers may be more inclined to select a handful of topics, providing more detail on the broader theme that their program is designated under, rather than the target audience list.
I believe that it is in important to keep in mind that while one subject may be very important to a local community, and therefore it's PEG station, that same subject (or audiences) may not even be on the radar at a different PEG station and not show up as a theme or topic. Also, one station could have very narrow themes established while another has none. Standardization is important but it is only as useful as the details entered by the producers on a regular basis.
I hope this feedback will be useful. Please let me know if there are specific topics that the group would like more information on.
Jacob
Jacob Ostrum
New Media & Programming Facilitator
Portland Community Media