E-Agriculture

TOPIC 1

Rama Rao Darapuneni
Rama Rao DarapuneniPandit Jayashankar Telangana State Agricultural University (PJTSAU), Hyderabad, India India

[quote="Lisa-Cespedes"] TOPIC 1 Populating a repository with resources and metadata: The quality versus quantity dilemma [/quote] Obviously quantity first. As we are trying to introduce elearning in agriculture the dilemma on either quality or quantity is far away. we are now trying to evolve a system to make agri faculty comfortable to use this medium and contribute. In a way we are a step behind and hence trying to learn from others experiences rama Rao, NAARM, India

Federico Sancho
Federico SanchoIICACosta Rica

[quote="Lisa-Cespedes"] TOPIC 1 Populating a repository with resources and metadata: The quality versus quantity dilemma [/quote] I must agree with RAMA, quality is always first that quantity. We live in a world of information abundance, and the major role for our organization is to select what is of good quality and what is not, so users will be closer to the best. That doesn´t mean it will happen, knowing some of the limitations of our stakeholders. Quality criteria for a Ag learning resource will be something to define and promote better. All the best, Federico Sancho IICA in Costa Rica

Vassilis Protonotarios
Vassilis ProtonotariosNEUROPUBLIC S.A.Greece

[quote="Lisa-Cespedes"] TOPIC 1 Populating a repository with resources and metadata: The quality versus quantity dilemma [/quote] I will have to agree with the rest participants of this conversation. I would go for the quality, hands down. Since we refer to educational material I am positive that quality is the first priority, and quantity is of secondary significance. I can see no use of populating a repository with resources poorly described with metadata. Of course, quality is a subjective term but if some basic guidelines are followed or if a quality assurance mechanism is deployed then we can be positive about the quality aspect of our educational resources.

Nikos Palavitsinis
Nikos PalavitsinisGreek Research & Technology NetworkGreece

Dear colleagues, Thank you for the first round of answers to our general topic posed here! Adressing your interest on this, allow me to provide some additional questions on this topic, that aim to provide different directions for our discussions. To facilitate the discussion, we would like to ask you to start replying by indicating the question number addressed (i.e. Q1), so that our colleagues that read you answers can easily identify the question you refer to. It's not obligatory to answer all of these, so feel free to choose the ones you want to answer. It would be nice, to answer with separate "posts" on each question. In this way, we will be able to follow up on your questions more easily. 1. Which metadata standard are you using in your institutional/project repository? 2. How many and which metadata elements do you require for a resource to be uploaded in the repository? 3. How do you measure quality of the metadata within your repository? Do you have any mechanisms (automated or not) in place, to ensure the completeness, correctness, etc. of metadata? 4. As a content creator, do you find it easy to provide metadata for the resources you create? What are your experiences so far? 5. Would you prefer a repository with hundreds of thousands of resources with the minimum metadata attached to them (title, description and keywords) or a repository with significantly less resources, described with a comprehensive set of metadata elements (covering educational aspects, format requirements, classification aspects, etc.)? Justify you choice. With kind regards, Nikos Palavitsinis

Vyacheslav Gadomski
Vyacheslav GadomskiVinnytsia National Africultural UniversityUkraine

[quote="Lisa-Cespedes"] TOPIC 1 Populating a repository with resources and metadata: The quality versus quantity dilemma [/quote] I think that on the early development of the resource the quality should be preferred. Then you can immediately increase the number of materials, but after experts learn to work with. Then the speed and quantity will grow themselves.

Lisa McLaughlin
Lisa McLaughlinInstitute for the Study of Knowledge Management in Education (ISKME)United States of America

Q1. Which metadata standard are you using in your institutional/project repository? For OER Commons (www.oercommons.org) we created our own metadata format and then created cross-walks to export our data in a range of metadata formats: Dublin Core LOM OER Recommender Lisa McLaughlin OER Commons Manager USA

Lisa McLaughlin
Lisa McLaughlinInstitute for the Study of Knowledge Management in Education (ISKME)United States of America

Q2. How many and which metadata elements do you require for a resource to be uploaded in the repository? On OER Commons we currently require the following 12 elements: Title URL Subject Learning Resource Type Media Format Educational Context (Grade Level) Abstract Keywords Institution Conditions of Use URL Conditions of Use Description Conditions of Use Buckets (our own groupings of open license types, exps: Share Only, Remix and Share) Lisa McLaughlin OER Commons Manager USA

Lisa McLaughlin
Lisa McLaughlinInstitute for the Study of Knowledge Management in Education (ISKME)United States of America

Q3. How do you measure quality of the metadata within your repository? Do you have any mechanisms (automated or not) in place, to ensure the completeness, correctness, etc. of metadata? At OER Commons, our minimum standard requires that material on our site contains the 12 elements outlined in Q2. We eyeball all data imports from harvesting to determine how comprehensive their metadata is. We also prioritize the import of collections known to be highly reputable. We use a python check script to ensure that automated harvests contain all the required elements but we also do a fair amount of manual curation. When a collection is of high quality or contains OER that is not easy to find elsewhere, we take the time to manually curate the items and add the missing data elements where and when we can find them. Users on our site are able to review items, which provides another quality control mechanism. When an item is poorly reviewed, we go over it to determine whether or not it should remain in the collection. Lisa McLaughlin OER Commons Manager [email protected]

Lisa McLaughlin
Lisa McLaughlinInstitute for the Study of Knowledge Management in Education (ISKME)United States of America

[quote="Lisa-Cespedes"] TOPIC 1 Populating a repository with resources and metadata: The quality versus quantity dilemma [/quote] It really depends on your end goal. At OER Commons we have a fairly large collection, currently around 30,000 items. We have a multi-faceted approach to dealing with this dilemma. One way we ensure quality is by developing and highlighting micro-sites that showcase content we believe is exceptional around a theme (such as our current micro-green initiative which will contain a significant collection of agricultural OER). At our size, however, we think its important to shift to a recommender system model for ensuring quality in our larger collection. Our users can currently rate and review items but we are working on developing models for driving micro-contributions from users who care deeply about particular content areas. Once enough user-contributed content evaluation data is in place, we will shift to a user interface that enables the content with the highest rankings to rise to the top. Ultimately, repositories require significant user-contributions to differentiate them from the wide range of collections out there. Lisa McLaughlin OER Commons Manager USA [email protected]

Salvador Sanchez-Alonso
Salvador Sanchez-AlonsoUniversidad de AlcalaSpain

Q1) A IEEE LOM application profile. Q2) 2 different levels of conformance exist. At the very basic one, only a few metadata fields but this is not recommended, so in the "regular" form, users should provide about 20-25 metadata elements of information. Q3) Expert reviews through a metadata quality workflow Q5) I couldn't tell just like that. But if completely necessary to choose, I would prefer to have just a few wel described and relevant to the topic of my interest. For huge repositories with no metadata I already have google :) Best regards, Salvador