Inspire is a fictional webservice for researchers to find blind-spots in the current global scientific body of knowledge. These blind-spots are uncharted question area’s that need answered by theoretic or evidence based research. Basically Inspire can show researchers what work needs to be done, and what work is already done. This makes research more efficient, more fun to find missing pieces of the puzzle together, cooperative and less competitive.
In January of the year 2008 I have written down this idea, that was related to the SURFshare program at my work at SURF. This ambitious idea could be used to create a pathway of the direction the program could head towards. It should made use of the existing information infrastructure, with Open Access publications as a free body of knowledge, but needed some extra additional components where the state-of-art back then seized to exist. The idea was to stimulate the creation of a free and open academic information infrastructure, and a grand vision could fuel this. This idea never made it, because it was meant too ambitious, to far beyond the grasp, and also it was just one use case, yet research activities contained so much more.
I have written this down and published this information so it might inspire you, otherwise it would be lost. The original document in Dutch can be found here. Below follows a revised translation.
Inspire is a webservice that serves the academic community. Its key function is to support researcher in the creative process, by finding blind-spots in the current body of knowledge. Inspire can support the scientist in different search tasks in a research work-flow, by reducing the information overload.
- Question spotter: Inspire helps the researcher at the start of research by finding out if the research question hasn’t been asked before. This makes research more efficient, by reducing work that already has been done. Or it shows what research can be done again to double check and evaluate the outcomes of previous research.
- Blind spotter: Inspire can help point out what questions haven’t been stated before, yet obvious that they need to be stated and researched to find the answer to.
- Solution pointer: Inspire provides support to look for solutions and methods that are available in other area’s of research. e.g. the idea of path finding technology of peer-to-peer networks in computer science were found in biology, inspired by trail-scent search strategies used by ants.
What are the necessary functionalities of inspire in order to work for you?
- Publication list builder: Inspire can generate a publication list of the researcher. (The more information you provide the better the results. Providing name variances, an ISNI and ORCID helps.)
- Knowledge profile: Inspire is able to create a knowledge profile. This is a ‘fingerprint’ or concept map of the knowledge and expertise of the researcher. This fingerprint can be used in other Inspire services, for example to measure the proximity of another researchers. This profile is created from extracting information from the publications in the CV.
- Related literature (domain related): Inspire is able to offer the researcher new and related literature available within his own knowledge domain. This is done by matching the fingerprint with the body of knowledge within his designated knowledge domain. The researcher is able to receive new literature specific to his interest at this moment in his academic career. This reduces the information overload of unrelated material, and reduces effort for searching himself to get a quick up-date of the state-of-the-art.
- Related literature (cross domain): Inspire is able to offer the researcher related literature outside his own knowledge domain. The reason to do this, is to get inspired by the things happening in other research domains, broadening the horizon and creating new research opportunities. This is done by matching the fingerprint with the body of knowledge outside his designated knowledge domain.
- People you know: Inspire is able to create a professional network based on people you already know. This is based on your peers (people you cited) and co-authors.
- Related people: Inspire is able to build your professional network. It shows people related to you according the proximity of the fingerprints / knowledge profiles. You can meet people you didn’t know they existed before, yet can be very useful to be inspired by.
According to Joost. G. Kircz (in Modularity: the next form of scientific information presentation? Journal of Documentation. vol.54. No. 2. March 1998. pp. 210-235) we can identify four user types for discovery of scientific information resources. The informed reader, the partially informed reader, the uninformed reader, and the non-reader.
The informed reader This is the reader who knows what he or she is looking for and is able to find his or her way in the literature quickly. … This type of reader is not interested in the general parts but wants quick, direct access to the specific description of the experiments or theory and the specific results. …
The partially informed reader This kind of reader is not conversant with the specific research as such, but is interested in the general aspects that might be of use for their own investigations. … This type of reader wants to know how a particular paper fits within the broader spectrum of their own research, what the relations are with other methods in the same field, and what the connections are with related fields. …
The uninformed reader This is the group of readers who want to learn something new. They are curiosity driven open-minded browsers, who hope to get a fresh idea from fields that are either unknown or of which they have only a rudimentary knowledge. Authors’ names do not ring a bell, prestige and fame are unknown, as is the jargon and the intricate details of the paper. This type of reader needs a clear statement of the goal and embedding of the work, the uniqueness or particularity of this work compared to others and so forth. This method of reading is typical in the exploratory phase of a research project, and demands contextualised and clearly written language.
The non-reader This category consists mainly of science administrators who want only to know if a researcher is active. Bibliographic information (and nowadays also often impact factors etc.) is normally sufficient. These readers look for a place where the essential record keeping data are given. In addition, this type of reader wants a general statement of the goal of the research and possible claims by the author, why the research was performed and how it fits in the larger body of research within the field; a clear distinction between the author’s own previous work and the new work warranting this new publication, and between their own and other peoples’ work.
Possible applications / scenario’s
In this Inspire project we mainly focus on the exploratory phase in the research process where the “uninformed reader” is active.
To be able to distinguish relevant information sources for the uninformed reader, a concept map fingerprint of his own knowledge domain has to be determined. This fingerprint can be made by linguistic methods like weighted keyword density mapping by scanning the full text papers in his CV. keywords in older research weigh less than newer research; because interests might shift over time.
Inspire calculates for all available information sources (papers, persons, projects) a concept map fingerprint. Then it calculates the ‘distances’ between concept map fingerprints.
Visualisation: the uninformed reader is able to vary the distances from his own concept map fingerprint. This way he is able to shift different related information resources in and out the viewpoint. In the first circle will contain his own publications, the second will contain directly related work, cited material and the work of coauthors, further away will lay information resources that are indirectly related, or from other knowledge domains.
Filters can be applied on the viewpoints, so that the uninformed reader sees only person, only papers, only projects, only research grants, only patents, and so forth. Within the different viewpoints also information resources can be narrowed down by using key-word search.
Annotations: the information resources can be annotated using social tagging. This tagging helps to create a folksonomy of a particular resource. This helps to improve the concept map fingerprint of that particular information resource.
Ratings: With a like or dislike the uninformed reader can give a higher or lower weight to the distance of an information resource. This influences the information visible in the differrent viewpoints.
Suggestion function: an API that automatically comes with suggestions of related work. for example a programmer can make a addon in popular text editors or blog systems, and use the API to come up with related information resources that are not far distanced from the lines of words that are being written.
Relations: This system is usful for the partially informed reader when the relations between information resources are presented in the different viewpoints.
Blind spotter: When the above applications are finished, it is possible to create a blind spotting viewpoint. This viewpoint, explained earlier, makes is easier to find research questions that have been questioned before, and that questions have not been asked before.
I hope you enjoyed reading. If you feel like leaving a comment about what triggered you to get inspired by reading this, please don’t hesitate.