These observations from the symposium participants can be assembled into a characterization of a possible future society. Indeed, the rapid evolution of information technology may dramatically change the purpose for libraries over the next 40 years. As Breeding pointed out in his initial remarks, we can expect to see at least two generations of changes in library institutions before 2050, driven by changes in society. A library will need to track societal changes, focus on the set of activities that provide the highest benefit, and adopt relevant technology.
There are multiple emerging societal changes that were referenced or discussed by the symposium participants and that can be used to understand possible futures. One significant change is the emergence of knowledge bases, with the ability of the system to answer natural language questions.
The emergence of global knowledge bases is being driven by the development of global networking. This enables any person to discover and access information resources. Similarly, this enables a library to
149 federate with other information repositories. An institutional library may become a component in a global library. The institutional library will be able to focus on a local area of expertise, while providing access to information supplied by other libraries. This implies that a library can contribute local knowledge to a global knowledge base and be recognized as the authoritative source for specific knowledge. This also implies that a library may federate with other institutions to provide access to a comprehensive set of knowledge. The result will be a federated global knowledge base that is accessible by all persons.
A second major change in society is being driven through commercialization of education and
commercialization of knowledge bases. Online courses, distance education, and Massive Open Online Courses (MOOCs; Lewin, 2012) are already providing learning opportunities that decouple education from formal (and physical) attendance at an institution. When knowledge can be accessed over a network, the differentiation between a commercial knowledge provider and a library will be blurred. A user may select a resource based on credentials (how authoritative is the source). Similarly, a student may select an education institution based on the perceived value of the graduation certificate. The freedom to choose source of knowledge and source of education will require libraries to focus strongly on the needs and expectations of their primary constituents.
A third major change in society is being driven by the emergence of Big Data. The analysis of data is now recognized as a fourth form of science (observation, experiment, simulation, data analysis). It is no longer sufficient to manage large data collections; the ability to analyze the data is equally important.
Very large data sets are difficult to move. Hence the ability to analyze data in place will become
necessary. As institutions accumulate massive collections, they will become the authoritative source for specific sets of knowledge generated by specific data analyses. A library will be strongly motivated to enable computation on its digital holdings, to ensure that the knowledge content can be extracted and is available for use by other researchers. This positions the library as the authoritative source for that knowledge content and ensures their relevance within a global knowledge repository.
A fourth major change is the development of personal knowledge sources and crowd sourcing.
Individuals will become the “domain experts” for their own personal knowledge bases, while also being able to participate in global knowledge initiatives. For example, each person will be able to monitor their own personal health, monitor the environment in which they live, and build a knowledge repository that represents the impact the environment has on them (pollutant ingestion), and their impact on the environment (carbon neutrality). The aggregation of this information through
epidemiological studies will be important for global response to societal health issues. The ability for individuals to participate in global knowledge generation initiatives (e.g., as citizen scientists) is equally important. A simple example is the participation in research by students. This validates the student’s learning, enables students to make an original discovery, and motivates students to pursue a specific educational initiative.
150 The emerging capabilities provided by information technology evolution are transforming the ways in which individuals can discover, access, and use information and knowledge. The new capabilities are driven by the ability to handle projects of a very large scale (petabytes of data, gigabyte/second networks, teraflop execution rates, global interconnectivity), and by the transition from information-based access to knowledge-information-based access. These observations can be combined into possible future visions for the next two generations of libraries:
1. Transition to an authoritative institutional information repository. When born digital content dominates, a search for relevant information can be based on the content of documents and data sets. A simple form exists today through Google, which can index the digital contents of a library. An authoritative information repository validates assertions about the information content that is indexed. The assertions may be governed by statements about the
completeness of the collection, or about the source of the content, or about the consistency of the material. Each library will be able to publish assertions about the content. A user can then map their expectations about the information content to the library assertions to select a relevant resource. The ability to assert the properties of authoritative data will become increasingly important when information is published by competing institutions and governments.
2. Transition to an authoritative global knowledge repository. The simplest example of knowledge-based access is the IBM Watson demonstration.72 Human cultural knowledge was captured in a database less than a terabyte in size. Natural language queries were turned into evaluation of about 200 relationships across the content, with a response time measured in tenths of a second. The system outperformed human cultural knowledge experts. Google, Microsoft, Apple, and Wolfram are pursuing similar systems to answer natural language questions (Giles, 2012). The Google knowledge base has 500 million entities, indexed with tens of thousands of different types or relationships. The Microsoft knowledge base has 350 million entities. The Wolfram Alpha computational knowledge engine supports 50,000 types of algorithms and equations on ten trillion pieces of curated data. What is the impact of knowledge manipulation at this scale on the future of libraries? If you can discover relevant information through a natural language query, will you need a librarian?
Libraries are on the path to evolve from repositories for books, to information repositories, and to knowledge repositories. An information repository, such as Google, indexes the content of each book.
Relevant information can be found based on combinations of keywords. A knowledge repository indexes relationships. Relevant information is then found through the intersection of responses across multiple types of relationships. Relationships may include:
- Logical / semantic relationships
72 http://www-03.ibm.com/innovation/us/watson/what-is-watson/index.html.
151 - Temporal / procedural relationships
- Spatial / structural relationships - Functional /algorithmic relationships - Organizational / group relationships - Familial / genetic relationships
- Epistemological / systemic relationships
By evaluating many types of relationships, the search space for a query can be expanded. By then looking for the intersections across the relationship results, the answer can be found. This was the basis for Watson’s success. A knowledge base not only analyzes relationships across the local content, but can also correlate information with external knowledge repositories. If current trends continue, a global knowledge base will emerge over the next 40 years that indexes human knowledge.