Magaly Bascones (JISC)- Metadata: implications for librarians and end user

Editor's Notes

• #2: The metadata issues are not new. Metadata has been always part of library world. We knew it as cataloguing for print collections, with electronic resources the role of the metadata has been exacerbated but not always understood and well manage. In electronic resources metadata has been for a long time treated as an add on and accessory rather than an essential part of the products.
• #3: Today, we want to do another step towards the understanding of the issues caused by bad quality metadata. Today, we want to bring the perspective of the end user and librarians. How bad metadata affect them despite the efforts of libraries to propose them the best service investing in content and sometimes in last generation discovery systems. We will highlight the main challenges presented by metadata, why this matters and the impact inadequate metadata is having, to finish with some recommendations for stakeholders.
• #4: The theme about why metadata is important and the issues around the quality of metadata are not new. All the actors of the supply chain have a part to play and are affected. Here I will summaries where the different actors are in order to give the context of our presentation. Lets start with Libraries. Libraries are looking for end to end workflows from acquisition to discovery of electronic resources in order to automate tasks and release staff time. They want to ensure user satisfaction and value for money. Purchasing consortia, they expect that electronic resources will be provided with “adequate” metadata by the content providers when purchased. However, experience shows that in many cases this is not the case. Knowledge Base Plus, we support libraries to workaround failures in supply and accessibility. We are suppliers of curated metadata. Our work have help to proof that the supply of good metadata is possible. We have also helped libraries to distinguish between the problems on the supply from content providers and the delivery from system vendors. Content providers: they provide metadata of variable quality and they do not always adhere to commonly accepted standards/recommendations. System Vendors: the increasing complexity of library information systems has led to new challenges in the management of metadata. Although these systems promise to reduce the effort in managing discovery metadata, current experience shows a more mixed picture. Standards and Recommendations are necessary in the metadata world. Recommendations as KBART, help content providers to deliver ingestible and enriched metadata more useful for system vendors. Also, support librarians to be able to work with their content. KBART is a readable friendly format.
• #5: What we know up to now: We know that Metadata issues are not new. We know that there are recommendations on how to do better for example KBART that are not always follow. We know where all the actors of the supply chain are positioned, their positive contribution, and for some of them their struggle. Based on our researches and our day to day experience, we have identified that these are the current challenges, and we will present them to you along side, with why this matters and the impact they are having.
• #6: To understand the issues we will talking today it is important to keep on mind the 2007 report about link resolvers and the performance of the supply chain. This report already 10 years old made the point that the poor provision of metadata was undermining the technology of link resolvers. The provision was and is responsibility of content providers. As a result of this report, KBART format was created. KBART is a format adapted to the needs of knowledge bases today and as before, the provision of metadata by content providers is plenty of deficiencies. Content providers may have adopted the KBART format but there are delays, lack of accuracy and lack of updates. Why? We know a little bit on this. The KBART standing committee, where I am/was co chair run survey and the findings numbered here include lack of knowledge, other priorities and lack of customer’s requests. In this case of libraries…we will come back to this latter
• #7: In 2013 to address the problems related to provision, Jisc KB+ started its activities. The amin was to provide libraries the opportunity to manage their electronic collections using a good set of metadata. The KB+ team has excelled in the task as metadata provider. The data not only available in KB+ but injected in the supply chain. As today, KB+ has addressed the issues related to provision within the metadata we curate. The delays doesn’t exist, the quality is good and KB+ is an appreciated tool by librarians and it is trusted. Trust is very important in the metadata world. We could think that that is!, all is solved but it is not the case
• #8: What we know up to now: We know that Metadata issues are not new. We know that there are recommendations on how to do better for example KBART that are not always follow. We know where all the actors of the supply chain are positioned, their positive contribution, and for some of them their struggle. Based on our researches and our day to day experience, we have identified that these are the current challenges, and we will present them to you along side, with why this matters and the impact they are having.
• #9: Let’s see some examples. I have based these example in real queries received by the KB+ team. A librarian had a subscription to a bundle using a consortia agreement with the content provider for access to an specific collection. The agreement specified that new titles should be included. KB+ team has added these titles in the correspondent title list in November 2017 but up to last week the end users of the library who reported the issues continue without finding these new titles (here a couple of examples). This will affect usage, without accurate usage it is very difficult to calculate cost per use or having a complete understanding of the entitlements. This is not an isolated case, we had received more similar queries affecting other collections.
• #10: I have another example of missing titles. It is a digitalised collection. In this case, a librarian said: “The collection contains over 450 titles, however the current target available in our link resolver only contains 45” The access to this collection is free to libraries and end users but this doesn’t mean it is cost free. The success of this digitalisation projects is on their usage. Again, in a time when money is tight, funders are more strict to allocate funding digitalisation projects and they want reassurance that the chosen collections are useful. If only 10% of the collection is discoverable, it doesn’t look right, does it? I have the hint that records with not identifiers haven’t been processed (this is not very good for an archive collection). Also for the titles that were displayed, not always were identified as part of the free collection. This could be a problem of delivery of metadata or a problem on the processing of the metadata by the system vendors.
• #11: When the provision is not an issue, other issues have surfaced. The performance of the system vendors products System vendors can’t hide anymore blaming the content providers provision There is an interesting piece of research from IEEE, Julie Zhu about the performance of system vendors. She analysed how IEEE data looks like in different vendors knowledge bases. She explained that IEEE when through a process of improvement of their metadata and KBART files delivery. She explained in her article the steps they took and the time they spend. Here her results that interest us today.
• #12: As you see the KBART 2 fields are the fields that make files more relevant as they include ebooks information, title history and access type. In this field, content providers shouldidentify and distinguish paid content and free content. Again, even though content providers (some of them), files completed with enriched and basic information, this information just gets lost, unused
• #13: What we know up to now: We know that Metadata issues are not new. We know that there are recommendations on how to do better for example KBART that are not always follow. We know where all the actors of the supply chain are positioned, their positive contribution, and for some of them their struggle. Based on our researches and our day to day experience, we have identified that these are the current challenges, and we will present them to you along side, with why this matters and the impact they are having.
• #14: One of the preferred methods of content providers to present their products are bundles. Agreements negotiated by purchasing consortia include collections of journals and other electronic resources. There is the believe that there is value for money when buying a collection instead of the individual subscriptions. Every renewal season, content providers advertise their proposal with phrases that imply that more content is being offer (with new titles for example, or because the terms and conditions allow more access). Customers (in this case librarians) understand that they are gaining value for money. All this is very good, it is fair to show a good product and that the terms and conditions are advantageous. However, we know that the bundles of items have a weak point : metadata. The reality is that, in many cases, the bundles are not delivered with the adequate metadata that will allow full access and discoverability by end user and, more importantly for our point, to allow librarians to compare usage and value. It is known, that the resources contained within a bundle may vary, either at the end of a fixed-term agreement, or even during the lifetime of agreement. When metadata describing the content of the bundle is not available, it is impossible to even understand what’s changed. This can make comparisons of value over time impossible. Given the already-tense situation regarding the cost of publications, this has the potential to lead to cancellations.
• #15: The following example is similar but I analyse it differently. In this case the titles seemed to have been deleted from the link resolver’s target. Again in this occasion, a librarian said: ….. Her institution had a subscription to a collection through a multi year agreement and she noticed that some journals were deleted and she was asking for our support clarify the entitlements. I realised that these journals were labelled as the new journals proposed by the content provider. I decided to see what was the situation for other subscribers of the same agreement. Were these journals discoverable at all? So I choose a group of subscribers (10), to the same collection, and I started clicking…For one of them, it was impossible for me to search their holdings so they do not appear on the results.
• #16: Well here the results. I have scored them based on the percentage of titles that were discoverable. Libraries use different system vendors that are connected to their own knowledge bases and link resolvers, so as expected we have different results. Some of the institutions used the same vendor (Vendor B) (these are institutions 2,3,4,6, and 9) but they have different results. The results depend on the targets they have activated. I have also added for information the cost of the individual institutional subscriptions, the cost of pay per view. The libraries belong to different sizes institutions (identified by the Jisc bands) as an indicators of resources available. Only one institution had 100% score. The lowest score is 26 % from “institution 4” where we found only 5 journals out of 19. For “institution 6”, it wasn’t clear where their access were coming from so I was uncertain to score them the same way. We do not know how much the results are influenced positively due to the librarians checking and correcting. Well, I know a bit, I know that Institution 1 expends time checking what their link resolver. They nearly got it. Only one missing. If they have improved their situation manually, it would indicate a hidden cost for libraries when buying bundles. This is just an example and the problem is scalable. How many more titles are affected and nobody know about it? How much institutions are paying for no discoverable items? How many disappointed end user are?, For end users, simply these titles are not there.
• #17: At the end of the day everything we do and everything we don’t is a question of priorities
• #18: This part will refer to the first slide we visit, the one about the story so far and where the different actors of the supply chain are. As mentioned, Last year, the KBART Standing Committee run a survey to identify the obstacles of adoption of the KBART recommendations within content providers. When we asked them “why they haven't used the KBART Phase 2” they replied: no time/other priorities and, no demand from customer base The part that stroke me the most was to learnt about the lack of customer requests. Our assumption was that librarians needed help to argue the benefits of good metadata in an effective way. To support this area we worked with Sero Consulting to produce a paper addressed to librarians. The aim was to elaborated a short document with few bullet points describing arguments that librarians could use during their conversations with content providers to make the customer requests easier. To prepare this paper there were interviews to electronic resources librarians, senior librarians and other library staff. In this case, the results were again surprising: Libraries can often work around the failures in supply and accessibility through their discovery layers, either manually managing lists or with the help of services such as Knowledge Base Plus but libraries are missing opportunities to improve the situation at the source during purchasing and negotiation. Internal priorities are contradictory depending the role of the librarian. In addition, no all librarians are confident on their knowledge of the metadata standards and recommendations that could help on the interaction with suppliers. Finally System Vendors, we have seen in our examples that the performance of the systems provided by the vendors is a mixed picture. For all the examples we have mentioned the librarians contacted the relevant vendor and the KB+ team did it too. The responses were slow and not all the problems were solved. If corrections were carried out, they happened case by case and not as a revision of their data processing. They operate globally so their focus is not just on the UK market. As a result, when metadata issues are reported by an individual institution in the UK, system vendors may not perceive this to be a priority issue.
• #19: Librarians should agreed their needs regarding metadata and use all their interaction points with suppliers (content providers and system vendors) to pass the message across. From Procurement process to conversations with suppliers reps. This can have good results, I am aware of at least one case (regarding eBooks) where a major vendor proved unable or unwilling to comply with what the library consortium viewed as basic standards in metadata and was therefore excluded from the procurement. Services as Knowledge Base Plus has helped to bring more information to the table and break barriers as the believe that accurate and timely metadata is not possible. It is! And this has been possible thanks to a committed team and the support of the community of users. It is a win-win situation, librarians will save time and at the same time it will increase the quality of the metadata they all use. Libraries have the power as customers to influence vendors. Libraries can develop the dialogue around the provision and processing of metadata, both pre- and post-procurement. System vendors need clarity about pain points and to plan changes in their development processes and resourcing. The library community can use existing user groups and can move things forward making sure that current customer’s experiences reach prospective customers. Content providers: Metadata is part of the product and it will influence the view of the customers. Reduction of usage could lead to cancellations. A working metadata supply system can only be achieved if stakeholders step in and see themselves as owners of the problem. Clearly, each actor has somewhat different priorities, and makes a different balance of costs and benefits, but high-quality metadata benefits all stakeholders in the supply and use of library content.
• #21: I did a similar presentation at UKSG that went really well and it was one the most attended breakout sessions. During the sessions we asked attendees to share their main worries related to metadata…47 people do it, using post its. An please do the same using your post its on the table, please give them to me at the end of the session. Thanks. Lets see some examples of what are the worries. Many are around , as said today the performance of system vendors. The impact of the current situation on libraries were also part of the worries we have. Bad metadata is a time consuming issue for librarians. And there were also some worries from the publishers present in the conference. They also have concerns and about the current situation
• #22: There is more, more or less, things I have mentioned today are known issues. There is something in the horizon that may change things and not necessarily positively and that can add another layer of complexity for the sector This is about KBART. Kbart as I said is a well known and used. Last year a new group has been formed, the KBART Automation. This group aims to delivered a set of recommendations to automatise the direct delivery of library holdings from content providers to system mvendors
• #23: The group has delivered their first draft. That is not public. Once they finish working on it the document will be public and open for comments. Basically, under these recommendations, content providers will give to each library a file witll their all accessible content including paid and free without identifying. Holdings will not very much complete. More or less what Elsevier is doing now
• #24: These are the benefits that have been stated by the first draft and the supporters of this initiative. I should say Exlibris is one of them. But there are some risks too…the risks come from me…