You are browsing the archive for Research.

Improving openness, transparency and reproducibility in scientific research

Jenny Molloy - October 24, 2014 in Guest Post, Reproducibility, Research, Tools

This is a guest post for Open Access Week by Sara Bowman of the Open Science Framework.

Understanding reproducibility in science

Reproducibility is fundamental to the advancement of science. Unless experiments and findings in the literature can be reproduced by others in the field, the improvement of scientific theory is hindered. Scholarly publications disseminate scientific findings, and the process of peer review ensures that methods and findings are scrutinized prior to publication. Yet, recent reports indicate that many published findings cannot be reproduced. Across domains, from organic chemistry ((Trevor Laird, “Editorial Reproducibility of Results” Organic Process Research and Development) to drug discovery (Asher Mullard, “Reliability of New Drug Target Claims Called Into Question” Nature Reviews Drug Development) to psychology (Meyer and Chabris, “Why Psychologists’ Food Fight Matters” Slate), scientists are discovering difficulties in replicating published results.

Various groups have tried to uncover why results are unreliable or what characteristics make studies less reproducible (see John Ioannidis’s “Why Most Published Research Findings Are False,” PLoS, for example). Still others look for ways to incentivize practices that promote accuracy in scientific publishing (see Nosek, Spies, and Motyl, “Scientific Utopia II: Restructuring Incentives and Practices to Promote Truth Over Publishability” Perspectives on Psychological Science). In all of these, the underlying theme is the need for transparency surrounding the research process – in order to learn more about what makes research reproducible, we must know more about how the research was conducted and how the analyses were performed. Data, code, and materials sharing can shed light on research design and analysis decisions that lead to reproducibility. Enabling and incentivizing these practices is the goal of The Open Science Framework, a free, open source web application built by the Center for Open Science.

The right tools for the job

The Open Science Framework (OSF) helps researchers manage their research workflow and enables data and materials sharing both with collaborators and with the public. The philosophy behind the OSF is to meet researchers where they are, while providing an easy means for opening up their research if it’s desired or the time is right. Any project hosted on the OSF is private to collaborators by default, but making the materials open to the public is accomplished with a simple click of a button.

Here, the project page for the Reproducibility Project: Cancer Biology demonstrates the many features of the Open Science Framework (OSF). Managing contributors, uploading files, keeping track of progress and providing context on a wiki, and accessing view and download statistics are all available through the project page.

Here, the project page for the Reproducibility Project: Cancer Biology demonstrates the many features of the Open Science Framework (OSF). Managing contributors, uploading files, keeping track of progress and providing context on a wiki, and accessing view and download statistics are all available through the project page.

Features of the OSF facilitate transparency and good scientific practice with minimal burden on the researcher. The OSF logs all actions by contributors and maintains full version control. Every time a new version of a file is uploaded to the OSF, the previous versions are maintained so that a user can always go back to an old revision. The OSF performs logging and maintains version control without the researcher ever having to think about it – no added steps to the workflow, no extra record-keeping to deal with.

The OSF integrates with other services (e.g., GitHub, Dataverse, and Dropbox) so that researchers continue to use the tools that are practical, helpful, and a part of the workflow, but gain value from the other features the OSF offers. An added benefit is in seeing materials from a variety of services next to each other – code on GitHub and files on Dropbox or AmazonS3 appear next to each other on the OSF – streamlining research and analysis processes and improving workflows.

 Each project, file, and user on the OSF has a persistent URL, making content citable. The project in this screenshot can be found at https://osf.io/tvyxz.

Each project, file, and user on the OSF has a persistent URL, making content citable. The project in this screenshot can be found at https://osf.io/tvyxz.

Other features of the OSF incentivize researchers to open up their data and materials. Each project, file, and user is given a globally unique identifier – making all materials citable and ensuring researchers get credit for their work. Once materials are publicly available, the authors can access statistics detailing the number of views and downloads of their materials, as well as geographic information about viewers. Additionally, the OSF applies the idea of “forks,” commonly used in open source software development, to scientific research. A user can create a fork of another project, to indicate that the new work builds on the forked project or was inspired by the forked project. A fork serves as a functional citation; as the network of forks grows, the interconnectedness of a body of research becomes apparent.

Openness and transparency about the scientific process informs the development of best practices for reproducible research. The OSF seeks both to enable that transparency, by taking care of “behind the scenes” logging and versioning without added burden on the researcher – and to improve overall efficiency for researchers and their daily workflows. By providing tools for researchers to easily adopt more open practices, the Center for Open Science and the OSF seek to improve openness, transparency, and – ultimately – reproducibility in scientific research.

Open Science at Tech4Dev 2014

Jenny Molloy - June 10, 2014 in External Meetings, Meetings, Research

Winner of the #LavauxContest photo competition at #t4d2014 from @GabrielaTejadaG

Winner of the #LavauxContest photo competition at #t4d2014 from @GabrielaTejadaG

Denisa Kera and Sachiko Hirosue pulled together a fabulous session at Tech4Dev 2014 #t4d2014 at the SwissTech Convention centre in sunny Lausanne. The conference was organised by CODEV and the UNESCO Chair in Sustainable Technologies and focused on ‘What is essential?’ in technology for development. Many answers to this question discussed throughout the three days converged around collaboration with communities, with many sessions highlighting examples of co-design and co-creation across a range of technologies for development including water, energy, healthcare and ICTs for education. This recognition of and commitment to participation and collaboration in reseach for development relates strongly to work completed by the open science working group and the OpenUCT initiative funded by IDRC (documented here, working paper available online).

The session ‘The Openness Paradigm: How Synergies Between Open Access, Open Data, Open Science, Open Source Hardware, Open Drug Discovery Approaches Support Development?’ covered a range of topics reflecting the breadth of practices that constitute open science but the two key areas of interest were open hardware for science and open data.

First speaker up was Professor Irfan Prijambada from Gadjah Mada University in Yogyakarta, Indonesia, who described the necessity of access to lab equipment for his microbiology research focused on agricultural practices and fermentation. Fermentation is important for alcoholic drinks but also the fermentation of cassava and rice to produce traditional Indonesian foods such as tapei. Further aspects of research in the Laboratory of Agricultural Microbiology centre around soil and water microbiology, including biodegradation and bioremediation in volcanic soils. As any microbiologist knows, the ability to observe small lifeforms and a sterile environment in which to culture and work with them are the two most essential research requirements in the lab. Prof Prijambada described the resultant difficulties of performing research effectively when dealing with obsolete and inadequate research equipment, relying on out of date microscopes with no digital image collection and plating microorganisms on agar in open spaces next to a bunsen burner with no access to a clean hood or laminar flow hood, both standard pieces of equipment for maintaining a near-sterile environment and ensuring samples are not contaminated. To add to these difficulties, applying for funding for equipment procurement at the university can mean a 12 month wait for processing and delivery even if the application is approved. There was a clear need for cheap, rapid and local supply of essential kit.

miCAM v3.2 on display at Tech4Dev. Photo by Jenny Molloy, all rights waived under CC0.

miCAM v3.2 on display at Tech4Dev. Photo by Jenny Molloy, all rights waived under CC0.

Step in Hackteria.org. In 2009 after a workshop in Yogyakarta run by Marc Dusseiller, an active maker and advocate of DIY biology and open source hardware, Prof Prijambada and his lab set about taking a DIY approach to lab hardware by creating their own clean hood and laminar flow hood, initially using a glassfibre filter but now employing a series of HEPA filters. The equipment was constructed in only 2 months for less than 10% of the cost of a commercial equivalent (1.2m IDR vs 15m IDR). Microscopes were constructed from webcams in less that one month costing 750k IDR instead of 7m IDR and were entirely adequate for research needs. Not only adequate, but aquisition of digital images allowed an automated colony counter to be developed. The importance and utility of these microscopes was explained by their developer Nur Akbar Arofatullah, a researcher at Gadjah Mada University, who founded the Lifepatch initiative and along with other hardware projects, has improved the DIY microscopes to the stage where a company is now offering a commercial version of the latest MiCAM v3.2 for those who don’t find DIY appealing. However, hands-on construction remains a key part of the educational aims of open hardware and Lifepatch are using the microscopes and their construction for a range of workshops pitched at different educational levels. Kindergarten students compare the width of their hair in Cyber Hair Wars, elementary students learn about plant and muscle cells, high school students construct their own microscopes while their teachers are taught how to run workshops themselves. University students are enthused with the DIY spirit and encouraged to apply these principles in their own education and research.

One area where Gadjah Mada University excels is community relations. Setting an example for all publicly funded research establishments, staff and students are expected and obliged to work with the community to achieve promotion within the university and there exists a dedicated Office of Research and Community Development. Within this ethos, DIY microscopes have been used to bridge knowledge between the university and community through workshops on sanitation and hygiene which make use of the microscopes and microbiology techniques to analyse water, take handswabs and analyse data on E. coli contamination. Lifepatch have run the Jogja River Project for several years, taking an integrative approach to water quality and river monitoring including participatory mapping and data collection on vegetation and animals all the way through to active clean up operations. Innovation in DIY hardware is rapid at Lifepatch and Gadjah Mada, with other projects including a vortex, rotator for incubating bacterial samples and a pipette stand. As per the example of MiCAM, the DIY approach is still compatible with commercialisation as people can buy pre-built hardware, thus offering the possibility of generating jobs and income but there are many questions around models for these activities which were of interest to the audience but could easily fill an entire session and were not covered in any depth (see reports here and here for an introduction).

In a related talk during a later session at a beautiful UNESCO World Heritage Site, the Lavaux vineyards on the banks of Lake Geneva, Dara Dotz presented on 3D printing open hardware during another session which touched on the creation of jobs and hyper local digital manufacturing capacity in Port au Prince, Haiti. Dara travelled to Haiti for three weeks and ended up staying for a year working for an NGO. She observed problems in water treatment plants and in hospitals which were caused by a lack of supplies and particularly spare parts due to a broken supply chain including long shipping and customs quarantine times, a culture of bribes and poor transport links for distribution. After a friend attended the delivery of 5 babies in one evening and had no option for tying off umbilical chords but using her own gloves, Dara realised that her background interest and contacts in 3D printing could be used to solve some of the issues of obtaining plastic parts and consumables. Having brought a Maker-bot 3D printer into Haiti, Dara trained a group of Haitians with basic education to use the printers and 3D design software and several potential uses were identified, with critical application being umbilical chord clips, splitters for oxygen tubing to allow multiple patients to receive oxygen from the same cyclinder and IV bag hooks to reduce the use of large IV stands which blocked space in already overcrowded wards.

3D Printing Umbilical Cord Clamps for babies in Haiti! from Not Impossible on Vimeo.

There were many considerations and design challenges to be addressed such as ensuring that designs addressed community needs and were designed with, by and for local people. In addition to empowering people to produce there own solutions to address real-time problems, the manufacturing method has the benefit of being on-demand, helping to ensure cleanliness of equipment, provides jobs and is also cheaper than importation of commercial equipment. Umbilical clips can be manufactured for $0.36 compared to $2.69 imported cost, representing a significant saving over time in this resource poor setting. Dara is now applying the same ideas to disaster zone supplies through the NGO fieldready.org and plans extensions to the Haiti project including importing CNC machines to allow manufacture of metal parts, creating a repository of designs for field supplies and increasing the use of recycled plastic waste for non-clinical devices and prototyping.

The Open Source Hardware approach advocated throughout this session is supported by the Open Source Hardware Association (OSHWA), a non-profit aiming to raise awareness of OSHW and to spur innovation by hobbyists, commercial and academic users. Gabriella Levine is President of the OSHWA Board and an artist with an interest in snake biomimicry. She introduced two projects designed for sensing water quality and clearing oil waste – Protei and Sneel, a snake biomimetic robot designed iteratively by Gabriella and documented online.

two sneels together playing from gabriella levine on Vimeo.

PROTEI PRESENTATION VIDEO from toni nottebohm on Vimeo.

These modular sailing and swimming robots allow sensors for oil, plastic waste,temperature, radioactivity and more to be attached and move through the water autonomously or via remote control, taking readings as they go. These concepts have been used in a range of water quality workshops and Gabriella runs hackdays exploring ideas around the design and deployment of water quality monitoring sensors and other hardware, including a water hackathon at Tech4Dev the following day!

With a variety of DIY and OSHW approaches and designs being prototyped and promoted in areas as important as sensors and even medical devices, a major question becomes how to ensure that quality is consistent and devices work accurately and safely. The current systems of quality assurance regulations in various countries are often either complex, expensive, time-consuming and a massive barrier to market entry – or non-existent. Kate Ettinger is working to develop a system for collecting information on quality and accuracy of OSHW projects in an open and transparent way using an open source hardware/software data collection system and an open data approach to making information available. This framework could apply to many projects but Kate used the examples of neonatal incubators and prosthetic limbs, with data being collected to accelerate responsive design and ensure ‘integrity by design’ throughout the development and deployment of open source medical devices.

OpenQRS in 30 Seconds from Kate Ettinger on Vimeo.

From open data for open hardware to open data as a research tool, Nanjira Sambuli from iHub in Nairobi described the use of crowd sourced data during the Kenyan elections in 2013 and contrasted data collected from Twitter and other social networks via passive crowdsourcing with active sourcing organised by Ushahidi. Conclusions presented were that machine learning algorithms are necessary to make collection of large datasets from high volume social networks viable and that there were surprising patterns and voices gathered through passive listening rather than active calls for information. Nanjira presented a framework developed by iHub for election data crowdsourcing emphasising the three V’s – viability, validity and verification.

jb_CrowdsourcingInfographice

Integrity and curation of scientific data was also highlighted in the final talk by Scott Edmunds of GigaScience , during which he described some excellent case studies of the power of openness. One example was increasing the rapidity of disease research during the E.coli outbreak in Europe in 2011, where BGI rapidly sequenced and released the genome as open data. The image of the chromosome map was later chosen as the front cover of a major report from The Royal Society in the UK on openness in science. Another example looked at the great scope for crowd sourcing the collection and analysis of open datasets. Research on Ash Die Back, an invasive tree disease, demonstrated several flavours of citizen science from publicly contributed geo-tagged photos of infected trees for OpenAshDB to gamification of genome data analysis via the Facebook game Fraxinus. It is also clear that citizens are very keen to support local research that is important for them and wish data to be made public to enrich their scientific and cultural heritage. The Puerto Rican “People’s Parrot” genome project took an endangered and much loved national symbol and sequenced its genome to learn more about its uniqueness and evolutionary history. This effort was funded by fashion shows, art projects, concerts, a branded beer and public donations. Scott focused on these successes but also discussed the challenges in increasing open data release, including ensuring researchers get appropriate credit and are incentivised to make their data available.

The BGI sequenced chromosome of the German E.coli outbreak strain.

The BGI sequenced chromosome of the German E.coli outbreak strain.

A common theme running through the presentations was that openness can be effective at accelerating innovation and enabling research in resource-poor settings. In addition, the scope for education and democratisation of the scientific process through involvement of local communities in scientific research and technological innovation has variously led to employment, empowerment and increased opportunities. The challenge now is to establish under what contexts this remains true and work to advocate and support open approaches where they can offer benefits for scientists and citizens in the global South. I hope members of this working group and the rest of the global open science community will be able to contribute to this mission!

What is content mining?

Ross Mounce - March 27, 2014 in Research

It’s simple really – you can break it down into it’s two constituent parts:

  1. Content

In this context, content can be text, numerical data, static images such as photographs, videos, audio, metadata or any digital information, and/or a combination of them all. It is a deliberately vague term, encompassing all types of information. Do not confuse content for the medium by which content is delivered; content is independent of medium in the digital sphere.

  1. Mining

In this context, mining refers to the large-scale of information extraction from your target content. If you extract information from just one or two items of content – that’s ‘data extraction’. But if you extract information from thousands of separate items of content – that’s ‘mining’.

content mining can involve multiple types of content!

It is important to emphasise that the phrase ‘text & data mining’ refers only to the mining a subset of the types of content one may wish to mine: text & data. Content mining is thus a more useful generic phrase that encompasses all the types of content one may wish to mine.

For my postdoc I’m using content mining to extract phylogenetic information and associated metadata from the academic literature. To be specific, the content I’m mining is text AND images.

The text of an academic paper contains much of the metadata about a phylogenetic analysis I want to extract, whilst the actual result of a phylogenetic analysis is unfortunately mostly only published as an image in papers (completely non-textual) – thus I need to leverage mining techniques for multiple content types.

Most mining projects tend to just mine one type of content, with text mining being perhaps the most common. An example question one might attempt to answer using text mining is: How many academic journal articles acknowledge funding support from the Wellcome Trust?

Peter Murray-Rust has many more different possible uses for content mining on his blog: http://blogs.ch.cam.ac.uk/pmr/2014/02/27/101-uses-for-content-mining/

In general, content mining is still an emerging and under-utilized technique in the research sphere – there is much work still to be done and billions of questions still to be answered. It isn’t the one-stop solution to everything, but for appropriate questions it can be the most powerful & comprehensive approach available. Get mining now!

Some suggested tutorials & resources you might want to start with:

10,000 #OpenScience Tweets

Open Science - March 20, 2014 in Media, Research, Tools

We have collected 10,000+ tweets using the #openscience hashtag on Twitter, and invite volunteers to help analyse the data. The twelve most-retweeted tweets are embedded below.

Happily, just over 4,600 accounts have participated in the Open Science community with its eponymous hashtag, in this span. The 10,000 tweets have accrued over ten weeks. Our own @openscience on Twitter has tweeted most, over 600 times at the hashtag, as well as having received the most retweets and @ mentions, over 8,000 in these 10,000.

We have modified the vis which came with the data via the satisfying TAGS effort shared by Martin Hawksey. We added looks at the numbers of mentions and of mentions per tweets for top tweeters, and rankings of top tweets for the past ten weeks to Martin’s default views. We will continue collecting tweets, but do note that in another month or so, we will reach Google Docs limits e.g. on numbers of cells. We will use additional sheets, so links to all data will have changed, just how depending on when you are reading this post. Ask us @openscience on Twitter.

Help wanted

More could be done; won’t you help? Leave a reply below or ping us @openscience on Twitter if you need edit access to the sheet itself but we would like to see data and analyses in other tools as well. Our work to this point is only to get something started.

Top #openscience tweets of the past ten weeks

 

 

 

 

 

 

The above list is not dynamic. The data collected and displayed here, however are dynamic and refresh themselves hourly.

Not all tweets which are about Open Science include the #openscience hashtag. In a perfectly semantic world, they would and when they can, they really should. It has helped to form a community among the 4,600+ accounts participating in these ten weeks and many others in recent years. A couple reasons the hashtag might not be used in a relevant tweet include the character limit on tweets and lack of awareness of hashtags or of the term Open Science.

We take our organising and leadership role seriously at @openscience on Twitter, an account shared by many in the community. We have a simple policy that all our tweets should be related to Open Science. Even at our account, not all our tweets include the #openscience hashtag, particularly as we discuss related concerns such as Citizen Science or Open Access. An example tweet from the time frame considered here, related to Open Science but not hashtagged as such is below. In this case, the limit on tweet length and the topic led to including #openaccess, not #openscience:

 

The most retweeted, Open Science related tweet of all time, so far as we know, did not use the #openscience hashtag but was lovely. From the Lord of Dance and Prince of Swimwear:

 

Building an archaeological project repository I: Open Science means Open Data

Michelle Brook - February 27, 2014 in Guest Post, Research

This is a guest post by Anthony Beck, Honorary fellow, and Dave Harrison, Research fellow, at the University of Leeds School of Computing.

In 2010 we authored a series of blog posts for the Open Knowledge Foundation subtitled ‘How open approaches can empower archaeologists’. These discussed the DART project, which is on the cusp of concluding.

The DART project collected large amounts of data, and as part of the project, we created a purpose-built data repository to catalogue this and make it available, using CKAN, the Open Knowledge Foundation’s open-source data catalogue and repository. Here we revisit the need for Open Science in the light of the DART project. In a subsequent post we’ll look at why, with so many repositories of different kinds, we felt that to do Open Science successfully we needed to roll our own.

Open data can change science

Open inquiry is at the heart of the scientific enterprise. Publication of scientific theories – and of the experimental and observational data on which they are based – permits others to identify errors, to support, reject or refine theories and to reuse data for further understanding and knowledge. Science’s powerful capacity for self-correction comes from this openness to scrutiny and challenge. (The Royal Society, Science as an open enterprise, 2012)

The Royal Society’s report Science as an open enterprise identifies how 21st century communication technologies are changing the ways in which scientists conduct, and society engages with, science. The report recognises that ‘open’ enquiry is pivotal for the success of science, both in research and in society. This goes beyond open access to publications (Open Access), to include access to data and other research outputs (Open Data), and the process by which data is turned into knowledge (Open Science).

The underlying rationale of Open Data is this: unfettered access to large amounts of ‘raw’ data enables patterns of re-use and knowledge creation that were previously impossible. The creation of a rich, openly accessible corpus of data introduces a range of data-mining and visualisation challenges, which require multi-disciplinary collaboration across domains (within and outside academia) if their potential is to be realised. An important step towards this is creating frameworks which allow data to be effectively accessed and re-used. The prize for succeeding is improved knowledge-led policy and practice that transforms communities, practitioners, science and society.

The need for such frameworks will be most acute in disciplines with large amounts of data, a range of approaches to analysing the data, and broad cross-disciplinary links – so it was inevitable that they would prove important for our project, Detection of Archaeological residues using Remote sensing Techniques (DART).

DART: data-driven archaeology

DART aimed is to develop analytical methods to differentiate archaeological sediments from non-archaeological strata, on the basis of remotely detected phenomena (e.g. resistivity, apparent dielectric permittivity, crop growth, thermal properties etc). The data collected by DART is of relevance to a broad range of different communities. Open Science was adopted with two aims:

  • to maximise the research impact by placing the project data and the processing algorithms into the public sphere;
  • to build a community of researchers and other end-users around the data so that collaboration, and by extension research value, can be enhanced.

‘Contrast dynamics’, the type of data provided by DART, is critical for policy makers and curatorial managers to assess both the state and the rate of change in heritage landscapes, and helps to address European Landscape Convention (ELC) commitments. Making the best use of the data, however, depends on openly accessible dynamic monitoring, along the lines of that developed for the Global Monitoring for Environment and Security (GMES) satellite constellations under development by the European Space Agency. What is required is an accessible framework which allows all this data to be integrated, processed and modelled in a timely manner.

It is critical that policy makers and curatorial managers are able to assess both the state and the rate of change in heritage landscapes. This need is wrapped up in national commitments to the European Landscape Convention (ELC). Making the best use of the data, however, depends on openly accessible dynamic monitoring, along similar lines to that proposed by the European Space Agency for the Global Monitoring for Environment and Security (GMES) satellite constellations. What is required is an accessible framework which allows all this data to be integrated, processed and modelled in a timely manner. The approaches developed in DART to improve the understanding and enhance the modelling of heritage contrast detection dynamics feeds directly into this long-term agenda.

Cross-disciplinary research and Open Science

Such approaches cannot be undertaken within a single domain of expertise. This vision can only be built by openly collaborating with other scientists and building on shared data, tools and techniques. Important developments will come from the GMES community, particularly from precision agriculture, soil science, and well documented data processing frameworks and services. At the same time, the information collected by projects like DART can be re-used easily by others. For example, DART data has been exploited by the Royal Agricultural University (RAU) for use in such applications as carbon sequestration in hedges, soil management, soil compaction and community mapping. Such openness also promotes collaboration: DART partners have been involved in a number of international grant proposals and have developed a longer term partnership with the RAU.

Open Science advocates opening access to data, and other scientific objects, at a much earlier stage in the research life-cycle than traditional approaches. Open Scientists argue that research synergy and serendipity occur through openly collaborating with other researchers (more eyes/minds looking at the problem). Of great importance is the fact that the scientific process itself is transparent and can be peer reviewed: as a result of exposing data and the processes by which these data are transformed into information, other researchers can replicate and validate the techniques. As a consequence, we believe that collaboration is enhanced and the boundaries between public, professional and amateur are blurred.

Challenges ahead for Open Science

Whilst DART has not achieved all its aims, it has made significant progress and has identified some barriers in achieving such open approaches. Key to this is the articulation of issues surrounding data-access (accreditation), licensing and ethics. Who gets access to data, when, and under what conditions, is a serious ethical issue for the heritage sector. These are obviously issues that need co-ordination through organisations like Research Councils UK with cross-cutting input from domain groups. The Arts and Humanities community produce data and outputs with pervasive social and ethical impact, and it is clearly important that they have a voice in these debates.

Content Mining: Scholarly Data Liberation Workshop

Jenny Molloy - December 14, 2013 in events, Oxford Open Science, Research, Tools

The November Oxford Open Science meeting brought over 20 researchers together for a ‘Content Mining: Scholarly Data Liberation Workshop’.

Iain Emsley and Peter Murray-Rust kicked off proceedings by presenting their work on mining Twitter and academic papers in chemistry and phylogenetics respectively.

Next we tried out web-based tools such as Tabula for extracting tables from PDF (we were fortunate enough to have Manuel Aristarán of Tabula joining us remotely via Skype) and ChemicalTagger for tagging and parsing experimental sections in chemistry articles.

OOS

We then got down to business with some hands-on extraction of species from HTML papers and mentions of books on Twitter using regular expressions. All code is open source so you are welcome and encouraged to play, fork and reuse!

Peter’s tutorial and code to extract species from papers can be found on bitbucket and the relevant software and command line tools have helpfully been bundled into a downloadable package. Iain has also documented his flask application for Twitter mining on github so have a go!

If this has whet your appetite for finding out more about content mining for your research and you’d like to ask for input or help or simply follow ongoing discussion then join our

open content mining mailing list


Some furry friends joined in the efforts - met Chuff the OKF Okapi and AMI the kangaroo

Some furry friends joined in the efforts – met Chuff the OKF Okapi and AMI the kangaroo

Open and transparent altmetrics for discovery

Peter Kraker - December 9, 2013 in Panton Fellowships, Research, Tools

6795008004_8046829553

by AG Cann

Altmetrics are a hot topic in scientific community right now. Classic citation-based indicators such as the impact factor are amended by alternative metrics generated from online platforms. Usage statistics (downloads, readership) are often employed, but links, likes and shares on the web and in social media are considered as well. The altmetrics promise, as laid out in the excellent manifesto, is that they assess impact quicker and on a broader scale.

The main focus of altmetrics at the moment is evaluation of scientific output. Examples are the article-level metrics in PLOS journals, and the Altmetric donut. ImpactStory has a slightly different focus, as it aims to evaluate the oeuvre of an author rather than an individual paper.

This is all good and well, but in my opinion, altmetrics have a huge potential for discovery that goes beyond rankings of top papers and researchers. A potential that is largely untapped so far.

How so? To answer this question, it is helpful to shed a little light on the history of citation indices.

Pathways through science

In 1955, Eugene Garfield created the Science Citation Index (SCI) which later went on to become the Web of Knowledge. His initial idea – next to measuring impact – was to record citations in a large index to create pathways through science. Thus one can link papers that are not linked by shared keywords. It makes a lot of sense: you can talk about the same thing using totally different terminology, especially when you are not in the same field. Furthermore, terminology has proven to be very fluent even in the same domain (Leydesdorff 1997). In 1973, Small and Marshakova realized – independently from each other – that co-citation is a measure of subject similarity and therefore can be used to map a scientific field.

Due to the fact that citations are considerably delayed, however, co-citation maps are often a look into the past and not a timely overview of a scientific field.

Altmetrics for discovery

In come altmetrics. Similarly to citations, they can create pathways through science. After all, a citation is nothing else but a link to another paper. With altmetrics, it is not so much which papers are often referenced together, but rather which papers are often accessed, read, or linked together. The main advantage of altmetrics, as with impact, is that they are much earlier available.

clickstream_map

Bollen et al. (2009): Clickstream Data Yields High-Resolution Maps of Science. PLOS One. DOI: 10.1371/journal.pone.0004803.

One of the efforts in this direction is the work of Bollen et al. (2009) on click-streams. Using the sequences of clicks to different journals, they create a map of science (see above).

In my PhD, I looked at the potential of readership statistics for knowledge domain visualizations. It turns out that co-readership is a good indicator for subject similarity. This allowed me to visualize the field of educational technology based on Mendeley readership data (see below). You can find the web visualization called Head Start here and the code here (username: anonymous, leave password blank).

headstart

http://labs.mendeley.com/headstart

Why we need open and transparent altmetrics

The evaluation of Head Start showed that the overview is indeed more timely than maps based on citations. It, however, also provided further evidence that altmetrics are prone to sample biases. In the visualization of educational technology, the computer science driven areas such as adaptive hypermedia are largely missing. Bollen and Van de Sompel (2008) reported the same problem when they compared rankings based on usage data to rankings based on the impact factor.

It is therefore important that altmetrics are transparent and reproducible, and that the underlying data is openly available. This is the only way to ensure that all possible biases can be understood.

As part of my Panton Fellowship, I will try to find datasets that satisfy these criteria. There are several examples of open bibliometric data, such as the Mendeley API, and figshare API that have adopted CC BY, but most of the usage data is not available publicly or cannot be redistributed. In my fellowship, I want to evaluate the goodness of fit of different open altmetrics data. Furthermore, I plan to create more knowledge domain visualizations such as the one above.

So if you know any good datasets please leave a comment below. Of course any other comments on the idea are much appreciated as well.

Open Scholar Foundation

Jenny Molloy - December 6, 2013 in Announcements, Guest Post, Reproducibility, Research, Tools

This is a guest post from Tobias Kuhn of the Open Scholar Foundation. Please comment below or contact him via the link above if you have any feedback on this initiative!

logo(2)

The goal of the Open Scholar Foundation is to improve the efficiency of scholarly communication by providing incentives for researchers to openly share their digital research artifacts, including manuscripts, data, protocols, source code, and lab notes.

The proposal of an “Open Scholar Foundation” was one of the winners of the 1K challenge of the Beyond the PDF conference. This was the task of the challenge:

What would you do with 1K that would significantly advance scholarly communication that does not involve building a new software tool?

The idea was to establish a committee that would certify researchers as “Open Scholars” according to given criteria. This was the original proposal:

I would set up a simple "Open Scholar Foundation" with a website, where researchers can submit proofs that they are "open scholars" by showing that they make their papers, data, metadata, protocols, source code, lab notes, etc. openly available. These requests are briefly reviewed, and if approved, the applicant officially becomes an "Open Scholar" and is entitled to show a banner "Certified Open Scholar 2013" on his/her website, presentation slides, etc. Additionally, there could be annual competitions to elect the "Open Scholar of the Year".

An alternative approach (perhaps more practical and promising) would be to provide a scorecard for researchers to calculate their “Open Scholar Score” on their own. There is an incomplete draft of such a scorecard in the github repo here.

In any case, his project should lead to an established and recognized foundation that motivates scholars to openly share their data and results. Being a certified Open Scholar should be something that increases one’s reputation and visibility, and should give a counterweight to possible benefits from keeping data and results secret. The criteria for Open Scholars should become more strict over time, as the number of “open-minded” scholars hopefully increases over the years. This should go on until, eventually, scholarly communication has fundamentally changed and does not require this special incentive anymore.

It is probably a good idea to use Mozilla Open Badges for these Open Scholar banners.

We are at the very beginning with this initiative. If you are interested in joining, get in touch with us! We are open to any kind of feedback and suggestions.

Open science & development goals: round up & the way forward

Jenny Molloy - September 23, 2013 in Collaborations, External Meetings, Meetings, Research

This is a post by the team at OpenUCT (post by Sarah Goodier, photos by Uvania Naidoo) and will soon be published on the OpenUCT blog.

Open science and development were the two key points that brought together a diverse group of over 20 scientists, methodological experts and researchers last week at the University of Cape Town. From the 12–13 September, these experts in their fields gathered for an IDRC OKFN-OpenUCT Open Science for Development workshop to scope possible research areas of open science for development. The focus was on research could be undertaken and to strengthen networks around this broad topic across Africa, Asia, Latin America and the Caribbean.

Day 1 involved discussions around opportunities and challenges for each of the regions represented as well as available resources that could be used and shared. By the end of the day, the group was starting to draw on these potential avenues for exploring open science for development to shape research questions.

Continuing from day 1’s discussions, day 2 focussed on framing these research questions around open science for development. These questions were discussed by breakout groups who selected the top four out of the multitude of those suggested. This selection was no easy task in such a mixed bag of broad, conceptual questions and focused practical questions – a clear indication that there are many potential interesting research questions.

day2_1

Four key questions emerged that were taken forward in further discussion:

  1. What value framework is a prerequisite for open science?
  2. How can open science support visibility and communication of science outside formal academic structures?
  3. How can open science create education?
  4. How can the economic and social value of open science be measured?

Projects that could help to answer these main questions were conceptualised and expanded upon. Some of the broad areas that the suggested projects could address included education, increased public involvement as well as the implication of open science on cost and building value. A regional focus for the suggested projects was thought to be best, largely due to financial and time limitations as well as co-ordination issues. The overarching IDRC-backed research programme will help to create and develop further synergies between any funded projects.

day2_2

As part of maintaining the momentum created during over the course of the workshop, staying connected and growing the network by bringing other people with diverse perspectives on board are key actions going forward. All of us walked away from this workshop with a greater appreciation for open science and an understanding that, although diverse, open science is united by many similar practises across regions.

We ended with more questions than answers at the end of the two days – just where you should be when you’re scoping possible research questions. What comes next is an OKFN working paper pulling together all the discussion threads, questions and resources raised over the two days, which will inform a call for research proposals for projects involving and investigating open science.

Watch this space as open science spreads across the map!

day2_3

Open science & development goals: shaping research questions

Jenny Molloy - September 13, 2013 in Collaborations, events, External Meetings, Guest Post, Meetings, Research

This is cross-posted from the OpenUCT blog.

What do we include in our definition of open science? And what is meant by development? Two key questions when you’re discussing open science for development, as we were yesterday on day one of the IDRC OKFN-OpenUCT Open Science for Development workshop.

Participants from Africa, Asia and Latin America and the Carribbean have gathered at the University of Cape Town in an attempt to map current open science activity in these regions, strengthen community linkages between actors and articulate a framework for a large-scale IDRC-funded research programme on open science. The scoping workshop aims to uncover research questions around how open approaches can contribute to development goals in different contexts in the global South. Contextualization of open approaches and the identification of their key similarities and differences is critical in helping us understand the needs and required frameworks of future research.

Several key themes, which generally provided more questions than answers, came up throughout a day packed of presentations, discussion and debate: strategic tensions, inequalities, global power dynamics, and the complexity of distilling common challenges (and opportunities) over large geographical areas. Some of the key strategic tensions identified include the balance between the “doing” of open science as opposed to researching it, as well as the tension between high quality research and capacity building at an implementation level. Both tensions are centred on inextricably linked components which are important in their own right. This brings up the question of where should the focus be? Where is it most relevant and important?

The issue of inequality and inclusivity also featured strongly in the discussions, particularly around citizen science – by involving people in the research process, you empower them before they are affected. But this begs the questions: How open should citizen science be? Who takes the initiative and sets goals? Who is allowed to participate and in what roles? With regard to knowledge, a small number of countries and corporate entities act as gatekeepers of the knowledge produced globally. How should this knowledge be made more accessible? Will open scientific approaches make dialogue and knowledge distribution more inclusive?

By the end of the first day’s discussion, the workshop had surfaced opportunities and challenges for each of the regions, but many questions still remain in terms of how to address the complex issues at hand and bring together the complex and disparate components of open scientific activity. Day two of the workshop will be focused on articulation of research problems, possible areas of activity and the structure of the envisioned research programme.

Join the discussion via Twitter via #OpenSciDev.

by SarahG (Pictures by Uvania Naidoo)