Tourismusverband Publication Rules
Zaenal Akbar, Jose Maria Garcia
STI Innsbruck, University of Innsbruck,
Technikerstraße 21a, 6020 Innsbruck, Austria
2014-03-11
Semantic Technology Institute Innsbruck /
STI INNSBTRUCK
Technikerstraße 21a
A – 6020 Innsbruck
Austria
1
Contents
1. Introduction
2. Rules Model
2.1. Content To Channel Mapping
2.2. Rules Implementation
2.3. Workflows
2.4. Content Transformations
2.5. Publication Scheduling
Appendix A. Analysis of Published Items in the Facebook
1. Hotels
2. Food and Drink Establishments
3. Events
4. Trips
5. Place of Interest
6. News
Appendix B. Analysis of Published Items in the Twitter
1. Hotels
2. Food and Drink Establishments
3. Events
4. Trips
5. Place of Interest
6. News
Appendix C. Analysis of Published Items in the YouTube
1. Hotels
2. Food and Drink Establishments
3. Events
4. Trips
5. Place of Interest
6. News
Appendix D. Analysis of Published Items in the Blog
1. Hotels
2. Food and Drink Establishments
3. Events
4. Trips
5. Place of Interest
6. News
1. Introduction
This document presents the publication rules for Tourismusverband Innsbruck (TVb). The main functions of publication rules are (i) to define which contents will be disseminated to which channels and (ii) to define how the disseminations should be performed, for example to define the order of the disseminations to several channels.
To obtain the publication rules for TVb, we have conducted analysis on the TVb’s website and TVb’s social media channels. The website will be used as the source of content (input channel) and the social media channels as target (output channels). The sources of this analysis are:
- TVb’s Website (
- TVb’s Facebook (
- TVb’s Twitter (
- TVb’s YouTube (
According to the TVb’s Redaktionsplan 2014, there are another social media channels to be considered as dissemination target:
- Google+
- Blog (
- tumblr
It seems that those new channels are still under preparation, as they are not officially linked to the main website and have very limited contents. In this document, from those 4 new channels, wewill present the analysis result for the blog only.
Another important inputs were obtained from the person who is responsible on managing the TVb’s social media channels. The inputs are concerning the publication timing, publication recurrence, including content transformation for a specific channel.
From the ontology definition analysis previously conducted, we have identified several relevant information categories for dissemination from TVb’s website as follows:
- Hotels
- Food and Drink Establishments
- Events
- Trips
- Place of Interest
- News
Based on those categories, the publication rules are then determined by answering the following 3 questions:
- (Existence) Can we find similar categories in an output channel? If yes, then there is a mapping between the categories to that channel.
- (Workflow) Is there any data dependency between a channel to other channels? If yes, then the publication to this channel must hold until the publication to the other channels have successfully performed.
- (Content Transformation) How a content has presented in a channel?. This question is intended to determine the required content transformation for each category to fit it into the channel’s requirements.
For text analysis, the messages were classified into 3 categories based on their size as follow:
- Short: below 150 characters
- Medium: between 150 and 500 characters
- Long: more than 500 characters
2. Rules Model
The answer for each question from the previous section will be explained in more detail in the following sub-sections.
2.1. Content To Channel Mapping
The mapping is representing a relation between the content categories and the available channels. If a similarity between the content of the website and the content of a channel is detected then we conclude there is a mapping between them.
Table 2.1 The Mapping of Content to Channel(s)
No. / Content / Facebook / Twitter / YouTube / Blog1 / Hotels / Yes / No / No / No
2 / Food and Drink Establishments / Yes / No / Yes / No
3 / Events / Yes / Yes / Yes / Yes
4 / Trips / Yes / Yes / Yes / No
5 / Place of Interest / Yes / Yes / Yes / No
6 / News / Yes / Yes / Yes / No
As shown at table 2.1, the mapping relations have been existed between all categories (except for Hotels and Food and Drink Establishments) to all channels.
It is important to mention that the mapping is only indicating the relations between content and channels. It is not representing which items from a content were disseminated to a channel, as it will be explained later in a sub-section related to the content transformations.
2.2. Rules Implementation
After the mapping between content to channels is obtained, the rules can be constructed from it. The basic format of a rule is: IF <content> THEN Publish(<content>, <channel>), meaning if a new <content> is detected in the website then it will be published to the <channel>.
Each row from the mapping represents one or more rules and can be implemented according to a specific rule language. In this document, we are going to use the Drools Rule Language[1] to show the rules implementation in Drools Rule Engine[2].
Table 2.2 A rule implementation
import info.innsbruck.model.content.Events;global info.innsbruck.model.channel.FacebookWall channelFacebookWall;
global info.innsbruck.model.channel.Twitter channelTwitter;
rule “Publish Events to Facebook”
when
i : Events()
then
insert(ItemToBePublished(i, channelFacebookWall);
end
Table 2.2 contains a rule to publish Events to the Facebook channel. While the Events are imported from the knowledge base (KB), the channelFacebookWall is the global name parameter declared in KB.
2.3. Workflows
Another important aspect needs to be considered during determining the publication rules is the workflow. If there is a data dependency between two channels, then we have to make sure the publications a common content to both channels are in correct order.
From our analysis, most of the published post messages on Twitter have an URL (link to the detail information) to the Facebook channel. So there is a strong evidencethat there is a dependency between Twitter and Facebook where a content should be published to Facebook first before published to Twitter.
Table 2.3 Rule implementation for the Twitter (has data dependency with Facebook)
rule “Publish Events to Twitter”when
i : Events()
f : ItemToBePublished(i, channelFacebookWall)
then
insert(ItemToBePublished(i, channelTwitter, f.url);
end
In table 2.3, the rule has a second condition to check if the Events have already been published to the channelFacebookWall. The rule consequence (to publish the Events to channelTwitter), will be executed only after the publication to channelFacebookWall has completed. In this case, the URL from the publication to Facebook is required to be included as a link in the tweet.
2.4. Content Transformations
The mapping between content to channels in sub section 2.1 does not indicate which content items are disseminated to each relevant channels. While a content could has rich items, a channel has specific requirements that have to follow. The process to transform the content items into an accepted format based on the channel requirements is called content transformation.
We strongly believe that the existing transformations requires a major improvement. Several transformations have been causing the lose of potentially important items. For example, the posts in the category of Hotels in Facebook channel are only Pictures (no Name, Location, etc.). A similar case happens to the posts in the category of Food and Drink Establishments.
The existing content transformations from Facebook, Twitter and YouTube channels are shown in appendix A, B, and C respectively. The analysis result from the new channel Blog is shown in appendix D.
We would like to point out several important notes from the administrator of the TVb’s social media channels which are related to the content transformation as follow:
- The contents on Facebook are mainly Images
- Sharing a post with a link is less popular than a post without a link
- Not all images on Facebook are available on the website
- Twitter is suitable to disseminate quick news
2.5. Publication Scheduling
Another important aspect in publication rules is how often a rulewill be executed from a specific input to a specific output channel. By considering this aspect, a scheduled publication can be arranged automatically. For example, if there is a new incoming Event, then it is most likely a better idea to send out reminder several times to certain channels automatically.
According to the administrator of TVb’s social media, regular posting to Facebook is 2 posts per day and the publication to Facebook is synchonized automatically with Twitter but sometimes an independent Tweet is also required.
Appendix A. Analysis of Published Items in the Facebook
Facebook is a social network channel that support various data including text messages, pictures, videos.
1. Hotels
Table A.1 List of Published Content Items
No. / Items / Type / Example1 / Picture / Image / A picture of Adlers Hotel
2. Food and Drink Establishments
Table A.2 List of Published Content Items
No. / Items / Type / Example1 / Picture / Image / A picture of Cafe Central
3. Events
Table A.3 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / Silvesterlauf Innsbruck
2 / Date / DateTime / 31.12.2013
3 / Time / DateTime / 16:15
4 / Place / Text (Short) / Innsbrucker Markplatz
5 / Description / Text (Medium) / The sporty end of the year ...
6 / More / URL /
7 / Picture / Image
4. Trips
Table A.4 List of Published Content Items
No. / Items / Type / Example1 / Place / Text (Short) / Das Sellraintal
2 / Description / Text (Medium) / Das Sellraintal ist mit über 25 ausgewiesenen Touren und ...
3 / More / URL /
4 / Picture / Image
5. Place of Interest
Table A.5 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / Innsbrucks Olympia SkiWorld
2 / Place / Text (Short) / Alps
3 / Description / Text (Medium) / Winter sports in the heart of ...
4 / More / URL /
5 / Picture / Image
6. News
Table A.6 List of Published Content Items
No. / Items / Type / Example1 / Title / Text (Short) / And the winner is ...
2 / Description / Text (Medium) / The winner of the Innsbruck-Trails has been ...
3 / More / URL
4 / Picture / Image
Appendix B. Analysis of Published Items in the Twitter
Twitter is a Microblogging channel where the size of submitted text message (tweet) is limited to maximum 140 characters. A post message could also contains an image.
1. Hotels
No data.
2. Food and Drink Establishments
No data.
3. Events
Table B.3 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / Innsbruck läuft
2 / Date / DateTime / 31.12
3 / Place / Text(Short) / Martplatz
4 / Description / Text (Short) / … sportlichen Jahresausklang am Innsbrucker Marktplatz. Wenn's dunkel..
5 / More / URL /
4. Trips
Table B.4 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / #Nachtskitouren rund
2 / Place / Text (Short) / im #Innsbruck
3 / Description / Text (Short) / ...alle Informationen zu einer schnell wachsenden 'Sportart'...
4 / More / URL /
5. Place of Interest
Table B.5 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / Innsbrucks Olympia SkiWorld
2 / Description / Text (Short) / Skifahren in der Hauptstadt der Alpen ...
3 / More / URL /
4 / Picture / Image URL /
6. News
Table B.6 List of Published Content Items
No. / Items / Type / Example1 / Description / Text (Short) / Innsbruck, 31-12-2013, 9:20, -5 C, sunny. Have a nice last day of the year
2 / More / URL /
Appendix C. Analysis of Published Items in the YouTube
YouTube is a video sharing channel, the registered users able to upload a video along with description. From the description, meta information were extracted such as as the location of a video was taken, user generated description, date, etc.
1. Hotels
No data.
2. Food and Drink Establishments
Table C.2 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / Adlers Bar Innsbruck
2 / Place / Text (Short) / 12th floor of the Pema-building, located next to the main station
3 / Offers / Text (Short) / Cool rhythms, Livemusik, cocktails, astounding views
4 / Description / Text (Medium) / Innsbruck is a very special location to experience the ...
3. Events
Table C.3 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / Innsbruck’s Christmas Markets
2 / Time / Text (Short) / About four weeks before Christmas
3 / Place / Text (Short) / Wiltener Platz, Old Town
4 / Description / Text (Medium) / Get enchanted by the flair of the Tyrolean Christmas Markets and the lovely smell of ...
4. Trips
Table C.4 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / Mieming’s idyllic trails
2 / Place / Text (Short) / In Mieming - one of Innsbruck’s Holiday villages
3 / Description / Text (Medium) / Enjoy a summer day among the mountains with your family ...
4 / More / URL /
5. Place of Interest
Table C.5 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / Moeserer See
2 / Place / Text (Short) / Few kilometers uphill from Telfs in the woods around Moesem
3 / Description / Text (Medium) / The lake in the midst of pine trees and a stunning alpine scenery is the warmest bathing lakes in Tyrol.
6. News
Table C.6 List of Published Content Items
No. / Items / Type / Example1 / Title / Text (Short) / Happy Birthday Alpenzoo Innsbruck!
2 / Description / Text (Medium) / The Alpenzoo in Innsbruck is celebrating it’s 50th anniversary this year. On this occasion a special issue stamp was produced ...
Appendix D. Analysis of Published Items in the Blog
A blog is a dissemination channel consisting of discrete entries (posts) displayed in chronological order. A post in a blog could cointains Text, Images, Videos and typically provides commentfeatures to its visitors.
In the time of this analysis was performed, the TVb’s Blog was still under preparation and had 2 posts only.
1. Hotels
No data.
2. Food and Drink Establishments
No data.
3. Events
Table D.3 List of Published Content Items
No. / Items / Type / Example1 / Name / Text (Short) / “ART Innsbruck” ist einzigARTig
2 / Date / Text (Short) / 20. - 23. Februar 2014
3 / Time / Text (Short) / Do. - Sa.: 11 - 20 Uhr
So.: 11 - 19 Uhr
4 / Place / Text (Short) / Innsbruck, Messehalle D+E
5 / Description / Text (Long) / Rund 70 Austeller aus zehn Nationen präsentieren auf einer Fläche von ca. 4000m² zeitgenössische ...
6 / Picture / Images
7 / Links / URL /
4. Trips
No data.
5. Place of Interest
No data.
6. News
No data.
1
[1] The Drools Rule Language,
[2]The Drools Rule Engine,