E-Agriculture

Introduction, Objectives, Background Note for Discussion and Summary of Week 1

Introduction, Objectives, Background Note for Discussion and Summary of Week 1

The international CIARD initiative (http://www.ciard.net) is working to make agricultural research information publicly available and accessible to all. Among its actions are advocating for and promoting open access, improving applicability and enabling effective use of data and information in agricultural research and innovation.

  Challenges and opportunities

Knowledge generated from  agricultural  research should be easily accessible and  so easily taken up  by  agricultural researchers  and the  agricultural  community at large. This issue is always more  vital in the face of rapidly  shifting  challenges such as climate change, transboundary  pests and  diseases, effectively use of water for agriculture, combating desertification and degradation of fragile soils and managing agricultural biodiversity.

Conventional pathways for communicating research outputs, through scientific publications (journals etc) and face to face events, have been enriched by new digital formats including the newer social media such as wikis, informal blogs and online communities of practice. The volume of  outputs of agricultural research  such  as  scholarly  or informal  publications, blogs, discussion forums, institutional directories  and very importantly raw data, is rapidly increasing.

Whereas an increasing amount of information becoming available represents an enormous opportunity, a challenge exists in the scattered way in which this information is being made available.  The  sources of information have become more heterogeneous  and  so coordination and coherence has become more problematic, making it more difficult to share data and information efficiently within and between scientific communities and more importantly with the wider range of stakeholders involved in agricultural innovation systems.

New approaches have to be adopted to facilitate sharing without resorting to tight
coordination and centralization, which, for political and logistical reasons, are not
foreseeable in the current spontaneous proliferation of data sets and services in agriculture.

This is where the current trends  in information technology and web practices represent a unique opportunity to develop a coherent framework for sharing agricultural information.

Approaches like Web 2.0, Linked Data, the OAI-PMH protocol, etc., are all means of
improving the interoperability1 of  distributed information and datasets without requiring strict coordination or the omologation of the software and data environment.

The CIARD partners are organizing a series of international consultations (through electronic and face to face events) to build a framework for action in making outputs of agricultural research truly accessible through coherent management, sharing and exchange of knowledge, information, and data. This background note is the basis for a discussion on the e-Agriculture platform from April 4 to 15.  The e-consultation  will be followed by an expert meeting  in Beijing in June 2011 with the same title. These two events will be an exercise in describing the current status and analyzing the needs for  tools, standards and infrastructures, leading on to defining future actions.

   Background Figures

Scientific publication and data production is growing at a much faster rate than ever before. Figure 1 charts the increase in numbers of articles indexed by MEDLINE from 1950 to 2010. The rate of increase in publication has clearly risen, especially since the year 20002, but the production of scientific papers is only the tip of the iceberg.

                     

 

Figure 1. MEDLINE-indexed articles published per year


Behind growing numbers of scholarly publications is a growing amount of scientific data. Furthermore, scholarly papers are no longer the  only  way in which scientific information is exchanged. Researchers are using more social platforms such as blogs3 to discuss results before they are published in scholarly journals or after they have been published.

Data on publication rates in agriculture are not readily available, but it is quite clear that the trend will be similar, although  perhaps  less dramatic, to the other  life sciences which are monitored by MEDLINE.

The second graph shows  that the  steeper  increases in publication  rates since 2000 is mainly due to the entrance of a few large countries with so-called "emerging economies",

                    

Figure 2. Thomson Reuters. Web of Science Database

 

especially China, Brazil and India, into the scientific mainstream. Taking 1990 as a base, Brazil has increased its scientific production by 800% and China by 1200%.
The  growth  of  these key  new players in the scholarly communication arena  has  made interoperability  a  more  important global issue, with special regard to the handling of languages other than English.

   Objectives of the e-Consultation

The CIARD partners aim to provide an opportunity to discuss some specific questions that will provide resource material for the expert consultation that will follows in Beijing. The econsultation has the following objectives:

  • To increase awareness about  opportunities  and technical options for creating  an infrastructure for data interoperability;
  • To create a list of discussion points about the current state of the art of sharing and interoperabilty in the agricultural research context as basis for the discussions of the expert consultation in Beijing;
  • To provide suggestions for actions that will enable increased sharing and exchange of data in agricultural research and innovation.

 


1 Interoperability http://en.wikipedia.org/wiki/Interoperability
2
http://altmetrics.org/manifesto/ Last accessed March 2011
3 See as an example ScienceBlogs
http://scienceblogs.com/ Last accessed March 2011

  

<p>What always come to my mind, which also was emphasized in the 2010 meetings with the prepared Road Map , is the relevance of <span style="">the research work to poor farmers. This off course should be reflected into the information to be shared. The problem which I feel important is how to ensure that the published research information has considered the needs of poor farmers and approved by farmers and other partners within the innovation system. This would remain as a missing link in the whole process, especially in many of the developing countries, even if emphasis is to be made in the future in the shifting of the research into more people-centered. Now, is it wise to think of &quot;<b>what are the proper mechanisms/standards that could help articulate, document and publish quality and relevant information or research outcomes?</b>&quot;. Or <b>How can </b></span><b>ICTs enable a more pluralistic, networked approach to information sharing</b><span style="">? The outcomes of the discussions could be some agreed guidelines to be useful to NARS and Regional Foras for further improvements in this regard. </span></p>
<!--[if gte mso 9]><xml>
<w:WordDocument>
<w:View>Normal</w:View>
<w:Zoom>0</w:Zoom>
<w:PunctuationKerning/>
<w:ValidateAgainstSchemas/>
<w:SaveIfXMLInvalid>false</w:SaveIfXMLInvalid>
<w:IgnoreMixedContent>false</w:IgnoreMixedContent>
<w:AlwaysShowPlaceholderText>false</w:AlwaysShowPlaceholderText>
<w:Compatibility>
<w:BreakWrappedTables/>
<w:SnapToGridInCell/>
<w:WrapTextWithPunct/>
<w:UseAsianBreakRules/>
<w:DontGrowAutofit/>
</w:Compatibility>
<w:BrowserLevel>MicrosoftInternetExplorer4</w:BrowserLevel>
</w:WordDocument>
</xml><![endif]--><!--[if gte mso 9]><xml>
<w:LatentStyles DefLockedState="false" LatentStyleCount="156">
</w:LatentStyles>
</xml><![endif]--><!--[if gte mso 10]>
<style>
/* Style Definitions */
table.MsoNormalTable
{mso-style-name:"جدول عادي";
mso-tstyle-rowband-size:0;
mso-tstyle-colband-size:0;
mso-style-noshow:yes;
mso-style-parent:"";
mso-padding-alt:0in 5.4pt 0in 5.4pt;
mso-para-margin:0in;
mso-para-margin-bottom:.0001pt;
mso-pagination:widow-orphan;
font-size:10.0pt;
font-family:"Times New Roman";
mso-ansi-language:#0400;
mso-fareast-language:#0400;
mso-bidi-language:#0400;}
</style>
<![endif]-->
<p>&nbsp;</p>

Mila M. Ramos
Mila M. RamosPhilAgriNetPhilippines

I agree.  Research products need to be translated into something that farmers can fully understand.  While many farmers are educated these days, technical language is not for them, not even for non-agriculturists.  While sharing info, it is very important that information is repackaged (or maybe simplified) for the intended audience. 

“建立全球农业数据/信息共享体系框架”电子磋商会简介材料(中文),包括背景介绍及各主题的基础材料。详见:

http://www.e-agriculture.org/sites/default/files/uploads/media/E-Confere...

(This is the background note in Chinese)

 

 

Valeria Pesce
Valeria PesceGlobal Forum on Agricultural Research and Innovation (GFAR)Italy
Week 1 hosted discussions on the first two questions posed in this forum: 1) What are we sharing and what needs to be shared? and 2) What are the prospects for interoperability in the future?
 
On the whole, in both threads participants approached the questions from different angles and introduced more and more aspects to the topics that were being discussed.
Some interesting issues that emerged in more than one post and permeated the whole discussion are:
 
1. The concept of "audience"
In the first thread many participants have argued that decisions on "what needs to be shared" depend on what the target audience is; some have illustrated the different patterns of "sharing" in the scientific community, others have highlighted the differences between the information needs of farmers and extensionists and researchers, and differences in access to technology and connectivity; some others have invited to also consider that when speaking about sharing and interoperability also other information systems (the machines) have to be considered as the "audience" of our information.
 
2. The "culture of sharing"
Many participants talked about the "fear of sharing" on the part of scientists and researchers   and the "struggle" for getting datasets and linked this to the issue of how agricultural research is rewarded. Others argued that sharing is often not done mostly because it causes extra work.
Among the proposed incentives to sharing are: making sharing part of one's work, providing proper mechanisms and pathways, demonstrating the benefits.
 
3. Considering all factors
Many participants invited to consider institutional and cultural factors, as well as networking and processes, alongside the technical factors ("Interoperability among people, concepts, data & systems"). Among these not-technical factors, some that were often mentioned were: considering / including all the stakeholders, considering the end user (linked to the issue of "audience" above), governance and how standards are set.
 
4. Prioritizing what needs to be shared or sharing as much as possible?
A number of posts clearly invited to prioritize what needs to be shared, depending on the audience, on information needs assessment procedures, on validation / peer-review etc. Other participants looked at the issue from another angle and broadened the range of what needs to be shared - also non-traditional and unconventional types of information / knowledge, even "tacit knowledge" - to the point of saying that everything should be shared and prioritization and selection can be done by the consumer, and to the "extreme" of proposing sharing individual "annotations" as triples.
 
5. The technical factors
Even if many users invited to consider other aspects and other factors, since the interoperability issue has a strong technical aspect, many posts tackled important technical issues, especially under Question n. 2:
 
5.1 Linked Open Data
Most participants agreed that LOD is the way to go for achieving interoperability
 
5.2 Technical requirements for effective sharing and re-use
It was stated or taken for granted in many posts that data need to be published in such a way that machine can process them (XML for instance, but XML does not express meaning, so RDF), that distributed datasets should refer to common published vocabularies / ontologies (preferably published as LOD) and that URIs should be used as much as possible: the utility of Linked Data derives from its use of URIs as “globally citable identifiers for making cross-references between things”.
 
5.3 Examples if interoperability mentioned in the posts:
- VIVO enables the discovery of research and scholarship across disciplines at that institution and beyond;
- eScienceNews automatically aggregates and semantically annotates contents from the web

Thank you Valeria, this summary is very useful for those of us trying to understand all the points raised in the discussion last week.

Many thanks Valeria for the focused summary. In point 4- Prioritizing what needs to be shared, isn't it possible that we emphasize sharing of best practices and success stories to encourage sharing of experience (methodology, data, information, and spirits) rather than data or information only.

Regards
Sallam

第一周的讨论主题是前两个问题:1)我们可共享什么,哪些需要共享?2)未来农业信息系统之间协同工作的前景如何?

整体来说,参与者从不同角度讨论了这两个问题,并且涉及的方面也越来越广泛。有几个问题不止一次的提到,而且贯穿整个讨论过程:

1.     “受众”的概念

在第一个问题中,很多人提到“共享什么”是由目标受众决定的:有人解释了科研领域的各种共享类型,有人提到了农民、信息推广人员、科研人员等不同的信息需求,以及他们在科技和网络连接等获取能力上的差异,还有人在提到共享和互操作问题时,认为信息系统(机器)也应该看做信息的“受众”。

2 关于“分享的文化”

很多参与者在谈到科研人员“害怕分享”、一些数据库不愿意开放链接等问题时,认为农业研究的回报问题是主要原因。还有人认为,分享有时太花时间反而会带来额外的负担。为了推动信息共享我们需要:把共享看做工作的一部分,形成合理的共享机制和共享路径,阐释共享的益处。

 

3 影响因素

我们应该将制度因素、文化因素、社会因素以及技术因素(人、概念、数据和系统之间的互操作)等都考虑在内。在非技术因素中,大家关注较多的是:要考虑各利益方的立场、考虑终端用户的需求和接受能力等(上面“受众”问题已提到)、管理问题以及标准如何设定。

4 哪些迫切需要共享或者需要尽可能共享?

很多人明确提出,共享内容的优先序是由受众、信息需求评估程序、有效性评估/同行评议等确定的,也有人从另外一个角度拓宽了共享的范围,他们认为非传统、非常规的信息/知识,甚至是“隐性知识”或者说一切信息都需要共享,而至于哪些迫切需要共享则由受众决定,我们所能做的则是把最基本的信息共享给受众。

 

5 技术因素

虽然影响数据/信息共享的因素很多,但是由于互操作性对技术性要求很高,因此很多人提到了技术因素,尤其是在第二个问题中:

5.1 链接开放数据

很多人认为链接开放数据是实现互操作的途径。

5.2 实现有效共享和再利用的技术要求

很多人认为甚至理所当然的认为数据应该以计算机可处理的形式发布(比如,可扩展标记语言XML,但是XML没有表达意义,因此相比来说资源描述框架RDF是更优选择),这样分散的数据集就可以使用通用的词汇或者本体(尤其是采用链接的开放数据),通用资源标识符(URIs)的使用也将更广泛,由此可链接的数据也将成为“事物之间相互参照的全球可引用标识”。

5.3 关于互操作的应用案例:

-VIVO实现了学术研究在机构内外的跨学科检索;

-eScienceNews可以自动加总和语义识别网页上的内容。