Jo Havemann presents a map containing more than 200 resources and supplementary data nodes across the spectrum of available tools, guidelines, events, and services by research discipline, also including general resources that are sortable by Open Science principle, language or country.
The post Guest Post – Mapping Open Science Resources from Around the World by Discipline appeared first on The Scholarly Kitchen.
Rebecca Lawrence discusses how connections across all aspects of the system are needed for open research to flourish and deliver upon its promise.
The post Guest Post — Why Interoperability Matters for Open Research – And More than Ever appeared first on The Scholarly Kitchen.
The DIAMAS project has just launched a survey which is of high relevance to the future of Open Access! DIAMAS supports Open Access Diamond and institutional publishing by setting new standards, […]
The post Help us understand Open Access Diamond and institutional publishing appeared first on SPARC Europe.
by André Vatter
The claim that AI language models are here to stay should be irrefutable by now. Although just introduced to the public in November 2022, ChatGPT has made rapid progress in this short time. How rapid? Let’s compare: After its founding, it took Twitter two years and Facebook ten months to build up a base of one million users. ChatGPT managed to reach this milestone in only five days. Two months after its launch, almost forty percent (German) of all Germans said they had heard of the chat robot or had already tried it.
But as impressive as the private adoption rate is today, what is more exciting is the escalations that ChatGPT has caused hitting the corporate world. Microsoft’s announcement of its plan to integrate the generative language model into its own search engine Bing caused sheer panic among the undisputed global industry leader Google. Google has been tinkering with an AI-supported web search for some time, but it has not yet been able to demonstrate that it is really ready for the market. There is a name, “Bard“, but CEO Sundar Pinchai is silent about concrete integrations. On the other hand, Microsoft was able to announce just a few days ago that its own search engine, which had hardly been noticed by users for decades, had to cope with a sudden rush of visitors:
“We have crossed 100M Daily Active Users of Bing. This is a surprisingly notable figure, and yet we are fully aware we remain a small, low, single digit share player. That said, it feels good to be at the dance!”
Redmond, Washington, is in an AI frenzy. In the future, there will hardly be a business area at Microsoft – whether B2B or B2C – in which ChatGPT does not play a role.
The disruption is also leaving its mark on the smaller competitors. Brave Search, the web search engine created by the US browser manufacturer Brave Software Inc. recently got a new AI feature. The “Summarizer” not only summarises facts directly at the top of the search results page, but also provides relevant content information for each result found. There are also changes at the privacy-focussed search engine DuckDuckGo, which has just launched “DuckAssist“. Depending on the question, the new AI feature taps Wikipedia for relevant information and offers concrete answers while still being on the search results page. But this is just the beginning: “This is the first in a series of generative AI-assisted features we hope to roll out in the coming months.”
All these integrations of AI language models into search engines are not about creating extensions to the existing, respective business model. It’s about a complete upheaval in the way we search the web today, how we interpret results and understand them.
Whereas the previous promise of search consisted of an effortful “I’ll show you where you might find the answer”, in combination with AI it suddenly advances to: “Here’s the answer.” Since their invention, search engines have only ever shown us possible ways where answers to our questions might be found. In fact, it has never been the inevitable goal of advertising-based business concepts to provide users with a quick answer. After all, the goal is to keep them in one’s ecosystem as long as possible in order to maximise the likelihood of ad clicks. This is also the reason why Google at some point began to present generally available information – such as times, weather, stock market prices, sports results or flight information – directly on the search results pages (SERPs), for example, in the so-called OneBox. The ultimate ambition is that no one leaves the Googleverse!
Intelligent chatbots, like ChatGPT, get around this problem. On the one hand, they transform the type of search by replacing keywords with questions. Soon, many users are likely to say goodbye to so-called “search terms” or even Boolean operators. Instead, they’ll learn to tweak their prompts more and more to make their communication with the machine more precise. And on the other hand, intelligent chatbots reduce the importance of the original sources; often there is no longer any reason to leave the conversation. Those who search with the help of AI want and get an answer. They do not want a card catalogue with shelf numbers.
We can already see that comfort does not come without critical implications. For example, with regard to the transparency of sources that we may no longer be able to see. Where do they come from? How were they selected? Are they trustworthy? Can I access them specifically? Especially in the scientific sector, reliable answers to these questions are indispensable. Other problems revolve around copyright. After all, AI does not create new information, but relies on the work of journalists who publish on the internet. How will they be remunerated if no one reads their texts and only rely on machine summaries?
Data protection concerns will not be long in coming. In communicating with the machine, a close relationship develops over time; the more it knows about us and can understand our perspective, the more accurately it can respond. In addition, the models need to be trained. Personalisation, however, inevitably means a critical wealth of data in the hands of third parties in return. In the hands of companies that will have to build entirely new business models around a question-answering game – quite a few of which, if not all, will be ad-supported.
AI provides answers. But not really to all questions at the moment. Search will change radically in a short time. Academic libraries with online services will also have to orient themselves accordingly and adapt. Perhaps the “catalogue” as a static directory or list will take a step back. Let’s imagine for a moment the scenario of an AI that has access to a gigantic corpus of Open Access texts. Researchers access several sources simultaneously, have them sorted, summarised in terms of content and classified: Have these papers been supported or falsified? The picture that emerges is of a new mechanism for making scientific knowledge accessible and comprehensible. Provided, of course, that the underlying content is openly accessible. From this perspective, too, here is once again a clear plea for Open Science.
So, how do academic libraries implement these technologies in the future? How do they create source transparency, how do they build trust and which disciplines of media literacy move to the foreground when new, machine-friendly communication is part of the research toolbox? Many questions, many uncertainties – but at the same time a great potential for the future supply of scientific information. A potential that libraries should use to actively shape the unstoppable change.
This might also interest you:
The post ChatGPT & Co.: When the Search Slot turns into an AI Chatbox first appeared on ZBW MediaTalk.
by André Vatter
The claim that AI language models are here to stay should be irrefutable by now. Although just introduced to the public in November 2022, ChatGPT has made rapid progress in this short time. How rapid? Let’s compare: After its founding, it took Twitter two years and Facebook ten months to build up a base of one million users. ChatGPT managed to reach this milestone in only five days. Two months after its launch, almost forty percent (German) of all Germans said they had heard of the chat robot or had already tried it.
But as impressive as the private adoption rate is today, what is more exciting is the escalations that ChatGPT has caused hitting the corporate world. Microsoft’s announcement of its plan to integrate the generative language model into its own search engine Bing caused sheer panic among the undisputed global industry leader Google. Google has been tinkering with an AI-supported web search for some time, but it has not yet been able to demonstrate that it is really ready for the market. There is a name, “Bard“, but CEO Sundar Pinchai is silent about concrete integrations. On the other hand, Microsoft was able to announce just a few days ago that its own search engine, which had hardly been noticed by users for decades, had to cope with a sudden rush of visitors:
“We have crossed 100M Daily Active Users of Bing. This is a surprisingly notable figure, and yet we are fully aware we remain a small, low, single digit share player. That said, it feels good to be at the dance!”
Redmond, Washington, is in an AI frenzy. In the future, there will hardly be a business area at Microsoft – whether B2B or B2C – in which ChatGPT does not play a role.
The disruption is also leaving its mark on the smaller competitors. Brave Search, the web search engine created by the US browser manufacturer Brave Software Inc. recently got a new AI feature. The “Summarizer” not only summarises facts directly at the top of the search results page, but also provides relevant content information for each result found. There are also changes at the privacy-focussed search engine DuckDuckGo, which has just launched “DuckAssist“. Depending on the question, the new AI feature taps Wikipedia for relevant information and offers concrete answers while still being on the search results page. But this is just the beginning: “This is the first in a series of generative AI-assisted features we hope to roll out in the coming months.”
All these integrations of AI language models into search engines are not about creating extensions to the existing, respective business model. It’s about a complete upheaval in the way we search the web today, how we interpret results and understand them.
Whereas the previous promise of search consisted of an effortful “I’ll show you where you might find the answer”, in combination with AI it suddenly advances to: “Here’s the answer.” Since their invention, search engines have only ever shown us possible ways where answers to our questions might be found. In fact, it has never been the inevitable goal of advertising-based business concepts to provide users with a quick answer. After all, the goal is to keep them in one’s ecosystem as long as possible in order to maximise the likelihood of ad clicks. This is also the reason why Google at some point began to present generally available information – such as times, weather, stock market prices, sports results or flight information – directly on the search results pages (SERPs), for example, in the so-called OneBox. The ultimate ambition is that no one leaves the Googleverse!
Intelligent chatbots, like ChatGPT, get around this problem. On the one hand, they transform the type of search by replacing keywords with questions. Soon, many users are likely to say goodbye to so-called “search terms” or even Boolean operators. Instead, they’ll learn to tweak their prompts more and more to make their communication with the machine more precise. And on the other hand, intelligent chatbots reduce the importance of the original sources; often there is no longer any reason to leave the conversation. Those who search with the help of AI want and get an answer. They do not want a card catalogue with shelf numbers.
We can already see that comfort does not come without critical implications. For example, with regard to the transparency of sources that we may no longer be able to see. Where do they come from? How were they selected? Are they trustworthy? Can I access them specifically? Especially in the scientific sector, reliable answers to these questions are indispensable. Other problems revolve around copyright. After all, AI does not create new information, but relies on the work of journalists who publish on the internet. How will they be remunerated if no one reads their texts and only rely on machine summaries?
Data protection concerns will not be long in coming. In communicating with the machine, a close relationship develops over time; the more it knows about us and can understand our perspective, the more accurately it can respond. In addition, the models need to be trained. Personalisation, however, inevitably means a critical wealth of data in the hands of third parties in return. In the hands of companies that will have to build entirely new business models around a question-answering game – quite a few of which, if not all, will be ad-supported.
AI provides answers. But not really to all questions at the moment. Search will change radically in a short time. Academic libraries with online services will also have to orient themselves accordingly and adapt. Perhaps the “catalogue” as a static directory or list will take a step back. Let’s imagine for a moment the scenario of an AI that has access to a gigantic corpus of Open Access texts. Researchers access several sources simultaneously, have them sorted, summarised in terms of content and classified: Have these papers been supported or falsified? The picture that emerges is of a new mechanism for making scientific knowledge accessible and comprehensible. Provided, of course, that the underlying content is openly accessible. From this perspective, too, here is once again a clear plea for Open Science.
So, how do academic libraries implement these technologies in the future? How do they create source transparency, how do they build trust and which disciplines of media literacy move to the foreground when new, machine-friendly communication is part of the research toolbox? Many questions, many uncertainties – but at the same time a great potential for the future supply of scientific information. A potential that libraries should use to actively shape the unstoppable change.
This text has been translated from German.
This might also interest you:
The post ChatGPT & Co.: When the Search Slot turns into an AI Chatbox first appeared on ZBW MediaTalk.
An interview with Kristin Hirschmann
Around ten years ago, the first Barcamp to focus especially on the topic of Open Educational Resources (OER) was held in Germany. For the anniversary event in October 2022 there were 446 registrations, 19 workshops, 56 Barcamp sessions, 235 minutes of video, keynotes and live podcasts ranging from Austria to New Zealand.
In this interview Kristin Hirschmann, project manager of the OERcamps organised by the J&K – Jöran und Konsorten training agency, reports on the development of the Barcamps and further OERcamp formats.
At the OERcamp in October 2022, the main focus was on the OER strategy of the federal government, introduced in the summer by Jens Brandenburg, parliamentary state secretary at the Federal Ministry of Education and Research (BMBF). The third day of the OERcamp was therefore completely dedicated to the strategy: Stakeholders of the OER community discussed and reviewed the different areas of activity of the strategy. We also made it possible for members of the OER community who were not able to be present at the OERcamp to comment on the results realised at the event. This commentary phase took place during the two weeks following the event via a collaborative document; 100 contributions were also added.
By agency J&K – Jöran und Konsorten for the OERcamp 2022 under CC BY 4.0
What’s more, after almost 3 years without face-to-face events, people felt it was important to bring the community together in one place. I received this feedback from many participants: the personal networking and exchange was even more important.
The first OERcamp took place at Bremen University in 2012. Since then, the OERcamp format has developed further according to current requirements: An OERcamp took place in Berlin every year until 2016. In 2017 there were then two innovations: the OERcamp was extended to four events for the first time: in the north, east, west and south. The OER Award was also established to select the best open educational materials in German-speaking countries. In 2018, the principle of one OERcamp in each of the four points of the compass in Germany was established. In 2019–2020, the OERcamp formats were refined so that the 5 aims of the OERcamp (qualification of OER, OER mainstreaming, networking and exchange, creation of specific materials and enabling a culture of sharing) could be implemented even more successfully. This means that alongside the participatory format of the Barcamp, there are now OERcamp workshops that focus on the creation and publication of OER as well as small, compact OERcamps that are affiliated to other events.
By agency J&K – Jöran und Konsorten for the OERcamp 2022 under CC BY 4.0
In 2020, the OERcamps were held in a virtual format owing to the coronavirus pandemic. With online formats such as the OERcamp webtalks or the OERcamp SummOERschool, we were able to respond to the urgent need of teachers to find, use and create digital teaching and learning materials quickly.
Further milestones that we were pleased about were the „Open Innovation Award 2020“(German) of the OE Awards Committee of “Open Education Global” at the 2020 OERcamp, the recognition of the work of the OERcamps with the 2020 EDUCAUSE Horizon Report the first international OERcamp in December 2021, OERcamp.global. This was, moreover, the first 48-hour festival on OER, and received 1063 registrations from 87 countries.
At the OERcamps, I am struck time and again by the participants, who are exceptionally enthusiastic. Of course, this comes on the one hand from the participatory format of the Barcamp. On the other hand, the formats of the OERcamp, such as the OERcamp workshop, are very needs-based and offer many more programmes than the participants can use: the participants therefore need the courage and openness to decide what they need for their personal OER journey from the abundance of programmes.
By agency J&K – Jöran und Konsorten for the OERcamp 2022 under CC BY 4.0
I also enjoy the great mix of people: those who are attending the OERcamp for the first time, and OER connoisseurs and pros. Experience has shown that around half the participants at the events are OERcamp newbies. The exchange that takes place there is very fruitful and also serves the aim of making OER accessible to a wider group of people.
In the OERcamp workshop, great focus is placed on the creation as well as the publication of OER. This is owing to the concept, because OER newbies receive a low-barrier start to working with OER thanks to the concrete utilisation of OER. This very specific learning-by-doing thereby enables them to experience the culture of sharing. I find this experience in particular a very important basis for continuing to engage with OER, and therefore also using the potentials of the OER community. The fact that we promote this active engagement with OER through the OERcamp format is something special, in my opinion.
The members of the OER community are not merely participants at an OERcamp who consume one-sidedly, but also central components who give input, ensure exchange and networking, thereby also ensuring the further development of OER and its dissemination.
From an international point of view, OER has even greater relevance than in German-speaking countries. The CC summit of Creative Commons, the Open Education Global Conference with the Open Education Week, the Open Education Policy Forum, the OER22 Conference, the Open Science Conference and the OpenEd are just a few events that focus on OER. At the events, it also becomes clear that OER plays a role on all continents. There is even more value placed on OER as a correct step towards more educational fairness and, above all, equity and social inclusion.
There is already a wide range of materials available for getting started in OER. This includes the #OERklärt video series (German), which explains the OER basics. This is published via OERinfo (German), the information centre for Open Educational Resources. The platform iRights.info (German) focuses on OER from a legal perspective. And the OERcamps themselves have published diverse materials that can also be further used. The campus of the OERcamps (German), for example, offers 12 online courses with know-how on OER. These include “100 great sources for OER”, “videos and audios as OER” or “online courses with and as OER”.
For librarians in particular, I recommend Fachstelle Öffentliche Bibliotheken NRW (German) – which emerged from the “oebib” blog, which has already been very active since 2015.
This text has been translated from German.
This might also interest you:
Kristin Hirschmann is a cultural and educational scientist. She works as project manager for the J&K – Jöran und Konsorten (German) training agency – a “think and do tank” for contemporary training. In this context, she designs and organises educational events for all educational fields and works in a content-related capacity on the topic of Open Education/Open Training.
Portrait: Kristin Hirschmann©
The post 10 Years of OERcamp: Community Get-together on Digital and Open Educational Resources first appeared on ZBW MediaTalk.
Author: Marcel LaFlamme, Open Research Manager, PLOS
When authors submit a Lab Protocol to PLOS ONE, they prepare a short manuscript that contextualizes their step-by-step protocol, describing the value it adds to the published literature and providing evidence that the protocol works. This additional context helps readers to decide whether and, if so, how to adapt the protocol for their own research.
In 2023, PLOS is making it easier for authors to share these protocol manuscripts as preprints, by expanding our partnership with the preprint server bioRxiv to include Lab Protocols.
During the submission process, Lab Protocol authors will now be asked if they want PLOS to forward their manuscript to bioRxiv to be considered for public posting within a few days. Facilitated posting to bioRxiv has been offered at PLOS ONE since 2018. Extending this service to Lab Protocols means that authors can share and get credit for their methods development work sooner, even as the peer review process unfolds.
In keeping with PLOS ONE’s multidisciplinary scope, Lab Protocol submissions are welcomed from any scientific field. Through the end of 2022, 61% of submitted Lab Protocols were in the life sciences, making bioRxiv an especially suitable partner to promote their early sharing.
“bioRxiv has always encouraged authors to post new methods as preprints,” said Richard Sever, co-founder of bioRxiv. “Formats like Lab Protocols that put protocols in context and provide data supporting their use are a great way to make methods more discoverable, and we’re delighted to partner with PLOS to make these manuscripts immediately available to researchers.”
The post Facilitated Preprint Posting is now available for Lab Protocols at PLOS ONE appeared first on EveryONE.
We are very excited to announce the publication of our 2022 Annual Report. Last year was an eventful year for SPARC Europe, and we were pleased to reconnect with our network […]
The post SPARC Europe highlights of 2022 appeared first on SPARC Europe.
OS Fair 2023 aims to bring together and empower open science communities and services; to identify common practices related to open science; to see what are the best synergies to deliver […]
The post Open Science Fair 2023 appeared first on SPARC Europe.
Open science is ushering in a new paradigm for research, one in which all researchers have unprecedented access to the full corpus of research for analysis, text and data mining, and […]
The post A joint strategy to strengthen the European Repository Network just launched appeared first on SPARC Europe.
Observations on reproducibility and research integrity from London STM Week
The post Research Integrity and Reproducibility are Two Aspects of the Same Underlying Issue – A Report from STM Week 2022 appeared first on The Scholarly Kitchen.
Iain Hrynaszkiewicz discusses PLOS’s Open Science Indicators initiatives and shares initial results.
The post Guest Post – How Do We Measure Success for Open Science? appeared first on The Scholarly Kitchen.
Did you know that over 60 European higher education institutions have policies covering the copyright of research publications and almost 45 institutions either have or will soon have rights retention policies […]
The post Project Retain Preliminary Findings: Rights Retention, Licensing & Copyright appeared first on SPARC Europe.
The Global Sustainability Coalition for Open Science Services (SCOSS), a part of SPARC Europe, has successfully supported three pledging rounds for Open Science Infrastructures (OSIs) helping them secure a sustainable future. […]
The post Going into fourth gear: SCOSS launches its 4th pledging round appeared first on SPARC Europe.
The 2023 LIBER Annual Conference will take place from the 5th – to the 7th of July at the Library and Information Centre of the Hungarian Academy of Sciences in Budapest. […]
The post LIBER 2023 Annual Conference appeared first on SPARC Europe.