Fostering Interdisciplinary Communication via Personal Knowledge Publishing

Esma Aïmeur 1, Gilles Brassard 2 Université de Montréal

C.P. 6128, Succursale Centre-Ville Montréal (Québec), H3C 3J7 Canada

Sébastien Paquet 3

NRC Institute for Information Technology 55 Crowley Farm Road

Moncton (NB), E1A 7R1 Canada

{aimeur,brassard,paquetse}@iro.umontreal.ca http://www.iro.umontreal.ca/~{aimeur,brassard,paquetse}

Final Revision, 26 December 2004

Abstract

The social and language barriers that exist between knowledge communities significantly hinder knowledge sharing and slow down progress. We investigate whether personal knowledge publishing, a web-based form of communication derived from weblogging, could facilitate the migration of knowledge across disciplinary boundaries. Case studies and the results of an online survey indicate that this approach is indeed helpful for interdisciplinary knowledge sharing. An important reason underlying this success is that this medium enables strong ties to develop between members of different knowledge communities. Moreover, the aggregate output of personal knowledge publishing can provide a useful foundation of represented knowledge on top of which to build intelligent systems.

Keywords: Interdisciplinary communication, Weblogs, Personal knowledge publishing, Social networks.

1 Supported in part by Canada’s NSERC. 2 Supported in part by Canada’s NSERC and the Canada Research Chair programme. 3 Work done while supported in part by scholarships from Canada’s NSERC and Quebec’s FQRNT.

2

Introduction

The rapid evolution of our world implies that learning and knowledge sharing are fast becoming a key challenge for individuals and organizations. The potential of online communities as environments for knowledge sharing has been recognized and explored in the literature on communities of practice. However, the questions of how online communities spring up, grow and stay healthy, and how to develop tools and structures that better foster, support and enhance knowledge sharing activities have yet to be adequately understood.

As knowledge advances, knowledge workers become increasingly specialized. Language barriers inevitably arise between areas of specialization: few people in a given community can readily understand discourses that originate from other communities. As researchers, we often have to solve problems for which a solution may be adapted from fields other than our own. But we find ourselves as tourists in foreign-language territory as soon as we undertake a moderately in-depth review of literature outside our area of specialization. Too often we reinvent the wheel, and find out only too late about it.

In this paper, we focus on the problem of sharing knowledge between distinct communities. As Swan remarks [7], innovation often requires a pulling together of knowledge and expertise from different parts of an organization. However, most hypermedia systems that support knowledge sharing are structured with an already existing group, community, or organization in mind as the fundamental starting point. When working in the context of such systems, people who are inquiring into an emerging issue that has not yet crystallized a group around itself run into a peculiar problem: they don’t have a “central place” where they can share ideas on that particular issue. In many cases, people likely to have an interest in it are typically scattered across existing communities. These observations are particularly visible in the world of scholarship, where strong social and language barriers exist between disciplinary communities.

This paper is concerned with the question of whether it makes sense to give the individual the opportunity to take a more central role in the dissemination of ideas in order to alleviate the problem of sharing knowledge across disciplines. The media of weblogging and personal knowledge publishing, which we describe below, have emerged in recent years as individual-centred community discussion support systems, which appear as interesting complements to journals, mailing lists, and other media for interdisciplinary-minded individuals.

Opportunities for Intelligent Systems Research

Knowledge elicitation has frequently been a key obstacle in the construction of successful intelligent systems. It is often arduous to get experts in some knowledge domain to make their knowledge explicit, especially at the frontier of the known. The difficulty is compounded when experts are required to provide formal representations of what they know. As intelligent systems usually need to build upon a large base of human knowledge, elicitation constitutes a bottleneck.

The work presented here highlights the ability of personal knowledge publishing to overcome this critical step of knowledge acquisition, by creating an environment in which large numbers of

3

experts are intrinsically rewarded for making their knowledge explicit and contributing it to a common pool. As such, it points to a new and promising potential basis towards the building of large-scale knowledge-based systems.

Moreover, it is possible to introduce more sophisticated and structured means of representation into personal knowledge publishing. An early example of this is the introduction of topic mapping, wherein users associate content with labels designating shared topical categories. An environment can thus be obtained in which knowledge is represented in increasingly structured ways, pushing both the Web and its users in the direction of the Semantic Web.

Systems like these could provide a rich, critical substrate for automated knowledge processing operations by intelligent systems, enabling them to classify content, answer queries, extract relevant information, and generate insights that would otherwise be much harder to obtain.

Another opportunity for intelligent systems research that resides in personal knowledge publishing is the unravelling of the networked structure of expert individuals. This is made possible by a large corpus of published personal knowledge. Since every piece of knowledge is directly tied to an individual, and because many of them comprise hyperlinks to pieces contributed by other individuals, mining the body of data could enable intelligent systems to perform tasks such as identifying authoritative experts in a given area.

Previous Work

Scholarly communication

The use of electronic channels for scholarly communication has been explored as far back as the seventies with the Electronic Information Exchange System (EIES), which enabled new journal forms and allowed knowledge bases to be developed for interdisciplinary professional groups [8, 9]. Making scholarly journal articles available on the Internet offers many benefits for the academic community, including reduced costs, instantaneous access, better ways of navigating the literature, wider and quicker dissemination, and higher impact of research [4].

Much attention has been devoted to the transition of articles to the electronic medium. However, as Harnad [2] observes, the process of knowledge development is largely contingent upon an informal prepublication phase in which researchers communicate about their ongoing work through telephone conversations, mail and email, and face-to-face encounters. Informal channels play an important role in facilitating awareness of new developments, as scholars often rely on personal networks for guidance in selecting material to look at [5].

Informal communication is crucial where interdisciplinary work is concerned, especially in the development of new directions in the space between disciplines, because these innovations don’t initially fit in any of the existing formal structures. The Internet obviously provides new ways to conduct informal communication, of which email is only the beginning. Harnad [2], extrapolating from the functioning of discussion lists, envisioned a transition of informal

4

communication to a vertical hierarchy of public, field-centric mailing lists, in which participant status increases and membership becomes more restrictive as one moves up the hierarchy.

As we will explain, personal knowledge publishing is more loosely structured, assuming the form of a network, which among other things allows growth through individual decisions without prior coordination.

Technology transfer

Efforts at enabling cross-pollination between knowledge communities predate the rise of computer-based information technology. In particular, this issue has been discussed in the context of technology transfer studies from the standpoint of organizational and behavioural dynamics [10]. In technology transfer, the issue is to ensure that innovations find their way to the people who can profitably use them. This can be facilitated along the vertical (research to development and to production) and horizontal (between production entities) axes.

However, this body of literature does not address interdisciplinary knowledge sharing within the level of research. Nevertheless, some of its more general observations are applicable, such as the role of barrier-crossing individuals [1], the structure of social networks and organizational culture [6].

Personal Knowledge Publishing

What we propose in this paper is for individual researchers to use personal knowledge publishing as a tool for building strong interdisciplinary research networks. These networks would enable them to locate relevant literature and experts outside their core research community much more easily. Although the tool in itself is not new, its application to interdisciplinary knowledge sharing has not been studied before. Personal knowledge publishing is a particular form of the practice known as weblogging. In order to explain it, we therefore first review the concept of a weblog.

Weblogs: Background and Definition

Weblogs may be viewed as an evolved form of personal Web pages, or “home pages”. The term, coined by Jorn Barger in 1997, refers to a website that is record of one’s journey across the World Wide Web. We shall use the term “weblog” to refer to a website that exhibits the following features, which are commonly agreed to be their defining characteristics:

1. Personal editorship. The content of the site is under the responsibility of a single person. 2. Hyperlinked post structure. The site’s contents consists in typically short posts that

usually feature hypertext links referencing material inside and outside the site. Each post is assigned a permanent hyperlink—or permalink—which is a URL that leads specifically to it.

5

3. Frequent updates, displayed in reverse chronological order. A weblog is a recurrent publication. The latest posts appear at the top of the weblog’s main page, and older content appears further down. New posts typically appear from several times per month to several times per day, which encourages repeat visits.

4. Free, public access to the content. The site’s content is freely accessible on the Web. 5. Archival. While older posts may disappear from the front page, they are archived and may

be accessed elsewhere on the site.

Brief history of weblogs

The very first weblog was World Wide Web inventor Tim Berners-Lee’s “What’s New?” page, which pointed to the new websites as they came online in the first years of the Web. Several new weblogs appeared around 1996. Although now collectively edited, Rob Malda’s Slashdot deserves mention, as it became phenomenally popular. In 1999, several free or inexpensive weblogging services were introduced. As using these systems requires little technical knowledge, the practice suddenly became much more accessible, and the number of weblogs has grown to millions in recent years, being adopted by a great diversity of people, young and old, homeless and billionaire alike.

One of the most significant things that happened with the recent growth of the weblog community is that weblogs have become an asynchronous conversational medium. Many editors use their weblogs to discuss things that have been said by another editor, using links to enable readers to follow threads. Arbitrary numbers of people can take part in such distributed, public conversations. Although the Web does not implement two-way linking, referrer logs—records of where visitors to one’s site come from—enable people to find who has been talking about what they have written.

Weblogs and other Internet forms of communication

While the medium of weblogs may appear at first glance like another incarnation of online many-to-many communication as previously implemented in the form of newsgroups, mailing lists and forums, its architecture results in crucial differences in terms of how participants’ attention is managed. Any member of a mailing list can write anything and it will automatically end up in front of every other participant’s eyes, which often results in information overload [3]—unless a moderator is introduced, which does not scale easily to large numbers of people. By contrast, since weblog contributions are posted on their author’s space, replying to someone’s post does not necessarily imply that the reply will be seen. Although weblog authors are free to say anything they please, they must earn the attention of readers, who allocate their attention on a per-person basis, by building a reputation over time.

6

Figure 1: Front page snapshot from the weblog of Lance Fortnow, a professor of computer science at the University of Chicago, taken on 8 November 2004.

Readers who have a weblog of their own also indicate interest with hyperlinks. The structure that results from the network of hyperlinks has a bearing on the quality of the material found in weblogs, since higher-quality content tends to be linked more and thus become more visible. Because authors engage their reputation in publicly linking to others, attempts to game the system in order to attract eyeballs—by techniques used only too often on the “standard” Web, such as link-farming—are likely to undermine one’s credibility and result in isolation.

Personal ownership and peer-to-peer communication patterns thus makes weblogs different from mailing lists and other collective discussion media. Weblogs may also be usefully contrasted with personal web pages. The key distinguishing characteristic of a weblog is that its main page is continually updated through the inclusion of new posts. This sets an expectation of updates with readers, which causes return visits and creates a stronger tie between author and reader than a static web page.

Personal Knowledge Publishing and its Dynamics

We build on the above definition of weblog to define personal knowledge publishing: an activity in which a knowledge worker or researcher makes her observations, ideas, insights, interrogations

7

and reactions to others’ writing publicly available in the form of a weblog. The content of a personal knowledge publication thus closely reflects the interests that her editor is researching.

Figure 1 provides an illustration of personal knowledge publishing. It represents the weblog of Lance Fortnow, who researches computational complexity. This site presents all the weblog features enumerated above, and more. The “About” box (1) shows that the site is written by a single person. One post is visible (2) on the snapshot (other posts exist below), featuring hyperlinks that go within and outside the weblog. The # sign seen to the right of the time-stamp of a post is a permalink that enables anyone to reference that post. Chronological structure is evidenced by the dates displayed in the page (3), which shows frequent updating of this weblog. Being accessible on the Web at URL http://fortnow.com/lance/complog, the site satisfies the free accessibility requirement (4). One of the links in the sidebar (5) gives access to archive pages for each week that weblog posts were written. Fortnow’s weblog additionally features subpages where readers may leave comments for the author and other readers (6). The author may reply to them in the same spaces, or in new weblog posts. (Private feedback is obviously possible using email.)

It should be noted that Fortnow’s weblog is an example of the “standard” use of personal knowledge publishing, that is, discussing a specialized topic of interest to its author. Fort-now is not applying the tool to the problem of sharing knowledge with people outside his discipline. Whether this tool can indeed serve to share knowledge across disciplines is not obvious from looking at most instances of personal knowledge publishing, and this is indeed the central issue investigated in this paper.

Implementation of Personal Knowledge Publishing

Several weblogging tools that have appeared over the last few years make personal knowledge publishing almost as technically easy as word processing. Weblog tools are typically implemented using a (server-side or client-side) database to store and organize posts and a Web-based interface to handle editing tasks. Publishing is usually done by way of File Transfer Protocol (FTP) to a Web server. Hosting and bandwidth may be provided by the weblog tool provider or by the editor herself.

How Personal Knowledge Publishing Works

How can a researcher build an interdisciplinary network around her personal interests using personal knowledge publishing as her principal means of communication? The process is comprised of four activities that we call exploring, pushing, pulling and regular reading. These activities are conducted in parallel.

8

Exploring

The exploration activity consists for the researcher in surveying the landscape of available content in weblogs to serendipitously find nuggets of particular interest to her. The best way to begin an exploration is probably to read a few weeks’ worth of content in a weblog by someone whom the researcher already knows and trusts to provide current and relevant knowledge, possibly on the basis of public reputation or prior interaction. This enables one to gather more links to interesting weblogs. An alternative way to begin an exploration phase is to use a weblog search engine (such as Feedster.com) or use a Web search engine for terms describing topics of interest, appending the words “weblog” or “blog” to the query.

The hyperlinked network structure of the space of weblogs makes a significant difference over websurfing in terms of extracting signal from noise. This is because once a reliable source has been found, it typically links to other regularly updated sources in the same neighbourhood of interests. As the most appreciated sources gather more inbound links, they are easily found. For the reader, each handpicked source subsequently acts as an information filter with very little noise. A form of authentication is enacted by the fact that the author of a weblog has exclusive writing rights and control over the site.

Pushing

The push activity is where the researcher comes out of the shadows. It consists in using one’s weblog to actively push information and knowledge out onto the Web. A researcher may push in a number of different ways, depending on the situation. She may link to and comment on a particular web page. She may simply link to a weblog post from another author that she finds relevant; she may complement the link with a quote from that post; and she may react by throwing in her own commentary. Another way to push is to write original posts or longer pieces. Yet another way is to add a link to someone else’s weblog in one’s blogrolling list. In all cases, pushing has the effect of giving a public indication of the researcher’s own interest in particular topics or people, enabling others to find her in the course of their own explorations.

Pulling

The pull activity is where the researcher exploits the responses that her weblog has generated to find yet more people. The basis for pulling is external links into one’s own weblog. Other weblog authors who find content interesting in one’s weblog will link to it from their own weblog. While the Web at large lacks bidirectional links, various methods exist to find out about the inbound links to one’s own weblog. Figure 2 illustrates one such tool, known as Technorati, which continually crawls over millions of weblogs and tracks the links between them. This enables it to give a current “citation report” for any weblog in its database. One way to enhance the value of pull is to post the questions one is asking oneself to one’s weblog. This often generates replies on other weblogs.

9

Figure 2: Technorati inbound link data for Sébastien Paquet’s Open Research weblog, captured on 15 November 2004.

Regular reading

Regular reading consists in looking up what interesting people have to say. Reading is valuable because it provides frequent exposure to new ideas that may support or complement the researcher’s interests, and because it allows one to maintain and strengthen relationships with other people with whom one would otherwise have no or little contact. Thus, it can serve to strengthen inter-community “weak ties” [6] and better exploit their natural tendency to generate new knowledge.

10

Weblogs and Traditional Scholarly Communication

We have already explained how weblogs differ from the public electronic media of forums and mailing lists. Here we contrast the medium of weblogs with the processes of public communication that are commonly used among scholars to communicate their findings, and explain how they are complementary.

Scholarly publishing and conference communications involve a formalized way of expressing ideas and a peer review process that are absent from personal knowledge publishing. Insofar as the peer review process is considered as the defining characteristic of scholarship, it is usually straightforward to distinguish between material that is scholarship and material that is not.

Personal knowledge publishing, on the other hand, is neither subject to formalization nor peer review. In this regard, it is closer to the hallway conversations, informal exchanges and spontaneous gatherings, in which scholars generate, examine and discuss ideas. Just as is the case with informal conversations, it can be difficult to determine which parts of them constitute scholarship per se and which ones don’t—and indeed, scholars usually don’t attempt it. However, there is no question among them that impromptu conversations are vital to generating and nurturing knowledge; and our work suggests that personal knowledge publishing can prove similarly useful in the creative process leading to the development of scholarly knowledge.

Before they are incorporated into the body of scholarship, fruitful ideas go through a life of their own. They often start out as hunches, vague connections and interrogations originating in a curious mind. From there, haphazard inquiry and informal conversations help them grow in strength. Some of these ideas provide the basis for more formal communications, which can gain further acceptance by being recognized through a peer review process.

As we explain in the rest of the paper, it is in the early stages of creative work that personal knowledge publishing can be the most helpful, by simultaneously giving its user the freedom to explore ideas of any nature and in any stage of maturity, and the possibility to document that exploration process and sharing the ideas with interested parties who may in turn contribute to growing and relating them with other ideas.

The usefulness of personal knowledge publishing does not lie in an ability to bypass or replace peer review; rather, it lies in its very openness to ideas that may well not stand past the early criticism of review committees. This is especially relevant in the context of interdisciplinary research, because the very notion of “peer” can be fuzzy, making peer review sometimes problematic in the absence of an established base, and in the case of conference submissions, when ideas can be defeated without the author having a chance to defend them. The overhead for posting ideas on weblogs is so small as compared to reviewed publishing that many more ideas can be given public exposure, circulated and tested. Gems are bound to emerge from the multitude, much as world chess champions usually emerge in Russia.

11

Case Studies

We describe the experience of two researchers who made use of the medium of personal knowledge publishing to exchange knowledge with people outside their discipline. The first is Lawrence Lessig, professor of Law at Stanford University. The second is Sébastien Paquet, one of the authors of this paper

Lawrence Lessig

Professor Lawrence Lessig is a world-renowned expert on constitutional law as it applies to cyberspace. Author of several books—see Fig. 3—he teaches and writes in the areas of constitutional law, contracts, comparative constitutional law and the law of cyberspace.

Lessig started a personal weblog1 of his own in August 2002. Over the following months, he used his weblog mainly to write posts in which he discussed political and legal ways to counteract excessively restrictive regulation on technology and copyright. Figure 3 shows a snapshot from Lessig’s weblog, taken on 15 November 2004.

Lessig used his weblog to interact with several other scholars who edit weblogs, among whom: Edward Felten, Professor of Computer Science and Director of the Security Inter-net Programming Laboratory at Princeton University, and editor of the Freedom to Tinker weblog2; Alex Golub, Ph.D. candidate in anthropology at the University of Chicago and editor of the Golublog3; Eugene Volokh, teacher of “free speech law, copyright law, and the law of government and religion” at UCLA Law School and co-editor of the Volokh Conspiracy weblog4; and David Weinberger, philosopher, writer, consultant and prominent thinker on the Internet and related issues, and editor of JOHO the Blog5.

With the exception of Eugene Volokh, the people listed above do not specialize in Lessig’s area. As Lessig’s interests lie at the intersection of law, technology, collaboration, innovation and even history, his weblog became a good platform to interact with people with different backgrounds who could contribute to advancing thinking on the themes he cared about. On several occasions, Lessig also used his weblog to request (and receive) help in preparing the Eldred case, in which he attempted to convince the American Supreme Court that Congress should not be allowed to indefinitely extend copyright terms.

1 http://www.lessig.org/blog 2 http://freedomtotinker.com 3 http://alex.golub.name/log 4 http://volokh.com 5 http://www.hyperorg.com/blogger

12

Figure 3: Front page snapshot from the weblog of Lawrence Lessig, a professor of constitutional law at Stanford University, taken on 15 November 2004.

Among the quantitative indicators of personal connections and knowledge exchange that have taken place because of Lessig’s publishing the Lessig Blog, weblog statistics service Technorati reported 1496 inbound weblogs providing 1919 inbound links6 as of 15 November 2004. More than two thousand public comments on posts have been left by visitors

Sébastien Paquet

Sébastien Paquet, one of the authors of this paper, started a personal knowledge publication7 called “Seb’s Open Research” in July 2002. In contrast to Lawrence Lessig, Paquet was hardly known outside his university when he began using the medium. He knew of only one researcher who shared his interest in personal forms of knowledge sharing. He otherwise had not had any contact with other people active in personal knowledge publishing.

He started off his weblog chiefly by citing posts from other weblogs and writing the thoughts that had occurred to him as he read them. At the end of July, he contributed the contents of one

6 http://www.technorati.com/cosmos/search.html?url=http://www.lessig.org/blog 7 http://seb.notlong.com

13

post on a discussion group8 and traffic started coming to his site twelve hours later, following a mention in David Carter-Tod’s Serious Instructional Technology publication9.

Tracing the new links pointing to his weblog that appeared in his referrer log enabled him to discover people who also wrote weblogs and had found his material. Many of those were unknown to him, but they shared his interest in new forms of knowledge sharing. Their backgrounds were very diverse, encompassing human factors, Web usability, information architecture, knowledge management, online learning, collective intelligence and complex systems theory.

On 3 October 2002, he made available online a 15-page research article titled “Personal knowledge publishing and its uses in research”10 and wrote a post on his weblog to introduce it. With no other particular effort, the article circulated appreciably, travelling from person to person over the weblog network and gathering links. Over the following months, it was mentioned by more than 90 different people, of which about half wrote comments of some sort on the content. Paquet was offered to republish the article on the Knowledge Board11 (the portal of the European KM community) a few months later, and translations were voluntarily prepared in French12 and in Polish13 by other weblog editors.

Paquet’s weblog also served as a jumping board to bring together people from diverse horizons with a shared interest in group-forming processes, resulting in the creation of a group-forming community site14. A close collaboration was borne out of this effort.

Among the quantitative indicators of personal connections and knowledge exchange that have taken place because of Paquet publishing Seb’s Open Research, Technorati reported 508 inbound weblogs providing 718 inbound links as of 15 November 2004—see Figure 2 again15; his comment monitor reports over one thousand public comments on posts in his weblog16. Prior to establishing his weblog, Paquet had almost no connections with people outside his institution. This effort thus proved very fruitful for helping him begin to carve out his place in the world of research.

8 http://groups.yahoo.com/group/klogs 9 Was http://instructionaltechnology.editthispage.com; reborn as http://itde.vccs.edu/blog/1 and http://david.carter-tod.com/wp 10 http://radio.weblogs.com/0110772/stories/2002/10/03/personalKnowledgePublishingAndItsUsesInResearch.html 11 http://www.knowledgeboard.com/cgi-bin/item.cgi?ap=1&id=96934 12 http://radio.weblogs.com/0105068/stories/2002/12/23/traductionDePersonalKnowledgePublishingAndItsUsesInResearch.html 13 http://www.gazeta-it.pl/zw/publikowanie_wiedzy_osobistej.html 14 http://aquameta.com/gf 15 http://www.technorati.com/cosmos/search.html?url=http://radio.weblogs.com/0110772 16 http://www.myelin.co.nz/commentmonitor/?blogname=Seb&usernum=0110772&url=http%3A%2F%2Fradiocomments.userland.com%2Fcomments&stype=rcs

14

Quantitative Survey Results

Over the months of December 2002 and January 2003, we conducted a survey in order to collect quantitative data on the usefulness of personal knowledge publishing for knowledge workers, and in particular for interdisciplinary communication. The survey was advertised through weblogs and a number of other online communities, and it was conducted online. Respondents visited a web page17 listing 25 questions on personal knowledge publishing. Several questions in the survey were included to collect data for other studies on personal publishing for knowledge management and are not directly used in this work. The questions relevant to our inquiry into inter-community knowledge sharing are shown in Table 1. The complete results of the survey are available on the Internet18.

In all, 177 people responded to the survey, but we take no account of the 6 who said they had never visited a weblog before (question #1). The results indicate that personal knowledge publishing, as a practice, allows knowledge workers to establish meaningful personal contacts with people in fields other than their own. The responses to questions #7, #16 and #17 are relevant to this issue.

Table 2 represents the distribution of answers to question #7. It is apparent from this data that trusting relationships can be established through the medium of weblogs, with 90% of all respondents answering positively to the question.

The next question that we consider examines the issue of interpersonal links across communities. Table 3 shows the distribution of answers to question #16. More than 80% of the respondents said they had been able to find people with a common interest but a different background, and 41% had found three or more such people, which indicates that weblogs enable the emergence of interpersonal links across communities of knowledge.

Table 1: Relevant questions in our survey on Weblogs. (Choices of answers in parentheses.)

1. Have you ever read a weblog? (Yes/No—If you answered “No”, you’re done!)

7. Did reading weblogs help you find people whose opinion you would be willing to trust on certain matters of professional interest? (No/Yes:1–2/Yes:3+)

16. Has your blogging activity helped you find people with professional interests close to yours but with a different professional background than yours? (No/Yes:1– 2/Yes:3–4/Yes:5+)

17. If you answered “Yes” to the previous question, do you think you could have connected with these people in another way than through weblogs? (No/Probably not/Probably yes/Definitely)

19. How often does weblogging help you fruitfully exchange ideas with people who specialize in a different area than your own? (Once a month/Several times a month/Several times a week/Several times a day)

17 http://www.blogstreet.com/surveys/weblog.html 18 http://www.iro.umontreal.ca/~paquetse/KnowledgeSharingSurvey.html

15

Table 2: Distribution of answers to question #7, “Did reading weblogs help you find people whose opinion you would be willing to trust on certain matters of professional interest?”

Count Percent Valid % Cum. % Valid Yes, >2 87 50.9 51.8 51.8 Yes, 1–2 64 37.4 38.1 89.9 No 17 9.9 10.1 100.0 Total 168 98.2 100.0 Missing 3 1.8 Total 171 100.0

Table 3: Distribution of answers to question #16, “Has your blogging activity helped you find people with professional interests close to yours but with a different professional background than yours?”

Count Percent Valid % Cum. % Valid Yes, >4 28 16.4 20.1 20.1 Yes, 3–4 29 17.0 20.9 41.0 Yes, 1–2 56 32.7 40.3 81.3 No 26 15.2 18.7 100.0 Total 139 81.3 100.0 Missing 32 18.7 Total 171 100.0

Table 4: Distribution of answers to question #19, “How often does weblogging help you fruitfully exchange ideas with people who specialize in a different area than your own?”

Count Percent Valid % Cum. % Valid Several times a day 2 1.2 1.4 1.4 Several times a week 17 9.9 12.1 13.6 Several times a month 66 38.6 47.1 60.7 Less than once a month 55 32.2 39.3 100.0 Total 140 81.9 100.0 Missing 31 18.1 Total 171 100.0

16

It is however possible that these links could have been established in other ways. Question #17 was designed to address this issue. Of the 113 respondents who answered ‘Yes’ to Question #16, more than 84% claimed that they would not (39/113) or probably not (56/113) have made such connections were it not for their involvement in a personal knowledge publishing network.

To summarize, the data collected in connection with these three questions lend support to the view that personal knowledge publishing allows knowledge workers to establish meaningful personal contacts with people in fields other than their own.

The survey results also suggest that personal knowledge publishing allows knowledge to be shared across fields. The responses to question #19 are relevant here. Table 4 shows the distribution of answers to this question. In the overall sample, 60% of respondents claim that knowledge exchanges with people outside their area occur several times a month or more. Judging from this distribution it seems reasonable to infer that personal knowledge publishing fosters better circulation of ideas across communities of knowledge.

Summary and Conclusion

Personal knowledge publishing is an activity whereby individual knowledge workers or researchers regularly update a website whose content consists in individual posts that feature hyperlinks and are presented in reverse chronological order. The website is public and its contents are freely accessible on the Web. Archives of old posts are kept in such a way that links to them (internal and external) still work after they have left the front page.

Personal knowledge publishing fosters interdisciplinary knowledge sharing in several ways. First, acting as a radically open conversational medium with low barriers to entry, it allows ideas to be exchanged and feedback to be obtained easily in a manner that is unconstrained by disciplinary boundaries, even when those ideas are not fully worked out or break with tradition.

Second, acting as an information filter, it enables content relating to an emerging interdisciplinary issue to be clustered in one place and made visible.

Third, it enables a web of trust to be built that is again unconstrained by disciplinary boundaries. Since the frequent-update characteristic is prone to induce repeat visits and conversations, persistent interpersonal trust relationships can be established and gradually strengthened among researchers. This enables the development of a new, practically noise-free (provided the weblogs are well chosen), social network where the linkage depends more on intellectual affinity than on membership in the same research community.

Fourth, the network can serve to efficiently assemble groups around emerging interdisciplinary ideas, because it implements a very efficient idea dissemination system. When a promising idea comes up, the word spreads quickly through the network. In a matter of days, it is possible for people who do not know one another, but happen to share a common interest, to find one another.

Two case studies and a survey conducted with users of this tool have confirmed its usefulness with respect to the goal of interdisciplinary knowledge sharing.

17

A number of research directions are being explored around the area of weblogs. Smart navigators, which take a user’s profile and lead her to the most relevant sites in a personal knowledge publishing network, would be desirable because exploration can be time-consuming, especially for new users. Appropriate standards for conversation mapping would be an asset. It is desirable to better bridge personal knowledge publishing and collective resources. Another enhancement would be to use personal user profiles to influence the display of links for each individual user.

Personal knowledge publishing is but one of the tools that we have considered for fostering interdisciplinary knowledge sharing. The use of Open Knowledge Shared Repositories, which is based on the concept of a wiki—a website that is editable and expandable by anyone19—will be developed in a companion paper.

Acknowledgements

We are grateful to VeerChand Bothra for his assistance in conducting the surveys and to Tommaso Toffoli for his advice and unfailing encouragements. We are also grateful to an anonymous reviewer as well as to the Editors of IEEE Intelligent Systems for their remarkably thorough comments on our original manuscript.

References

[1] E. Garnsey and S. M. Wright, “Technical innovation and organizational opportunity”, International Journal of Technology Management 5(3):267–291, 1990.

[2] S. Harnad, “Scholarly skywriting and the prepublication continuum of scientific inquiry”, Psychological Science 1:342–343, 1990.

[3] S. R. Hiltz and M. Turoff, “Structuring computer mediated communication systems to avoid information overload”, Communications of the ACM 28(7):680–689, 1985.

[4] S. Hitchcock, A. Woukeu, T. Brody, L. Carr, W. Hall and S. Harnad, Evaluating Citebase, an open access web-based citation-ranked search and impact discovery service, Technical report, Department of Electronics and Computer Science, University of Southampton, 2002.

[5] A. M. Odlyzko, “The rapid evolution of scholarly communication”, Learned Publishing 15(2):7–19, 2002.

[6] M. Ruef, “Strong ties, weak ties, and islands: Structural and cultural predictors of organiza-tional innovation”, Industrial and Corporate Change 11:427–449, 2002.

19 http://en.wikipedia.org/wiki/WikiWiki

18

[7] J. Swan, “Knowledge management in action: Integrating knowledge across communities”, Proceedings of the 34th Hawaii International Conference on System Sciences, page 7017, 2001.

[8] M. Turoff and S. R. Hiltz, “The electronic journal: A progress report”, Journal of the Ameri-can Society for Information Science 33(4):195–202, 1982.

[9] M. Turoff, S. R. Hiltz, A. N. F. Bahgat and A. Rana, “Distributed group support systems”, MIS Quarterly 17(4):399–417, 1993.

[10] F. Williams and D. V. Gibson (editors), Technology Transfer: A Communication Perspec-tive, Sage Publication, 1990.