From where the disruptive ideas for IT developments or why startups should not neglect scientific knowledge undertake

Scientific articles are important

Usually at the initial stage of a startup the author needs to study the large volume of scientific articles, to purchase strong theoretical base that will allow to select reference points both at problem definition, and by search of the solution. The correct approach to problem solving – to rely on the systematic, qualitatively executed researches which results can be published or recognized as a dissertation research. Quantitative data about behavior of users are very important for systems which promise significant improvement of UX. Examples it is possible to give a set here – from the normal master thesis of 2003 about users of Accenture "Social Software as a Source of Information in the Workplace"^[1]in which reveals as the staff of Accenture company uses social networks in the work, prior to numerous researches of the recognized specialist Steve Vittaker with measurements of efficiency of different user applications, answering such questions as: what percent of recently opened files users can find what depth of folders is optimal, etc.

The behavior of users is not always obvious

Personal subjective experience can differ from results of statistical analysis very strongly. Besides, it is necessary to have an opportunity to carry out comparison of data from several different systems. We see manifestation of behavior of the user in the existing systems while far from perfect (that is confirmed by a large number of separate estimates of that, what is the time users spend for information search, analysis of e-mail, etc.), and users can not realize shortcomings of the existing systems even. It is a problem of historical development: before a system did not become for professional users, and understanding of how such system should look at some point was gone. As a result now there are no systems intended for the solution of difficult tasks as, for example, writing of the thesis – the existing systems force the user to unjustified time expenditure. Meanwhile, ideal behavior and the technical solution corresponding to it it is possible to see and describe, only relying on wide experience and the theory. The lack of experience and ignorance of the theory leads to implementation of inefficient solutions.

For example, experiments showed that in the real environment it is inconvenient to users to use tags (them difficult to rename, organize, etc.). And at some point tags were one of the most widespread tools of the organization of information. In serious scientific literature and materials of conferences there are very valuable ideas about what features of a system are objective advantage: time for transaction, work speed, percent of successful transactions. The interesting moment that there is no uniform theory of the user interaction though in computer science there is a certain area studying this subject.

• Simple example. The idea about importance of integration of mail, files, etc. project data sources was sorted in 2006 in the report at the The Project Fragmentation Problem in Personal Information Management CHI 2006 conference^[2]. It was tested using Project Folders prototype.

• Difficult example. Douglas Engelbart's works where it describes the CoDIAK model (CONCURRENTLY Developing, Integrating, Applying Knowledge) for work with information. The competence of Engelbart who is the author of the modern concept of interactive computers is not called in question. In the 60th years he brought up a question of into what transactions it is possible to separate effectively work of the users who are busied with information. Also he created the Augment system which successfully showed the efficiency, including in aircraft industry (McDonnell Douglas company). Many modern IT specialists consider Engelbart the guru, and derive inspiration for development of own systems from systems (such as Augment and NLS) created by him.

Knowledge = happy users

So, the lack of znaniyevy base excludes a possibility of innovations (without preliminary researches the efficiency of innovations will be extremely low – more correct, normal for the venture industry through which go most IT-Innovations, but in space or nuclear sector such percent of success would be considered as catastrophic), the lack of innovations does not allow to solve a problem effectively (for the solution of new problems new approaches as old, obviously, have limited potential are required), the inefficient solution does not attract users. Besides, gradual completion of the project is improbable as there are no persons interested to pay for badly worked product, so and to pay back current expenses on development it will not turn out. Investments into it are also improbable since the project has neither brand, nor the qualified command, nor the innovative solutions, nor the customer base.

In the majority of IT startups founders believe that enough technical skills of programming and minimum idea of requirements of the market to create a successful product. But I claim that in IT there is scientific research and it is important!

In IT there is this science

What means "important" and why? The matter is that though computer science in Russia is a little known under this name, after all it is computer science – i.e., science about computers. In addition to mathematics and information science which occur first of all computer science includes cybernetics, cognitive sciences, sciences about communications, including between people, and many other spheres. In Higher School of Economics there is a faculty of computer sciences, but it is one of the few examples of comprehensive study of computer science. For the rest future developers finish mekhmat or faculty of applied information science that, naturally, does not give rather deep understanding of sections of this science and communications between these sections.

Since the 1960s scientists work on theoretical and practical questions of development of a qualitative product. Fans, how many serious professionals are engaged in development of computer technologies not so much. Many of them do career not of developers, and academic scientists. It is difficult to tell what should be ideal balance between science, practice and a dream, but, for example, importance of geological science and also its narrower areas for the oil and gas industry will hardly come to someone to mind to deny. Layers, breeds, evolution of Earth and many other questions are of not the just academic interest: engineering and finance solutions really depend on it. Therefore cannot but please that the situation with studying of work of users improves – in two almost passed decades of the 21st century number of annual publications in scientific magazines in the direction, for example, PIM (Personal Information Management) grew from scratch to more than seven hundred.

Let's take, for example, article of 2006 "The Project Fragmentation Problem in Personal Information Management"^[3] from the CHI 2006 conference (Computer-Human Interface). Since 2006 similar conferences are held annually and at them important scientific and practical results of the industry are presented. In the mentioned article in 2006 the question of fragmentation of project information between different formats (links, files, mail, tasks, contacts) is brought up. In this article there are results of specific experiments and if to study other works of authors of article, then it is possible to find also other important details of a problem, interrelation with the adjacent directions of scientific search and quantitative results of other experiments.

It is possible to give still a set of examples and links to scientific research, but the main thesis it is clear and so – without theoretical knowledge it is impossible to create an innovation which would provide competitive advantage. Yes, there are also other sources of competitive advantages, but for serious success serious preparation and the disruptive ideas is necessary.

Card of computer sciences

Click to open the complete version of the card in the PDF format

Sources of theoretical knowledge for development of computer systems

From where does such theoretical knowledge undertake? Where do teach them? Certainly, in higher education institutions on specialty computer science, it gives much more, than just ability to write a program code. Students, undergraduates and graduate students in computer science study an algorithm theory, data structures, bases of operating systems, computing complexity, the theory of programming languages, machine learning and many other things, not to mention the applied mathematics required for this purpose.

The specialty "programmer" appeared in our country in the early fifties. But just those who got the first vocational education in this area had the main impact on development of ADP equipment in the USSR. One of such pioneers of theoretical and system programming - Andrey Petrovich Yershov. Having gained the diploma of the programmer and having protected on a mekhmata of MSU in 1957, he headed department of theoretical programming of Computer center of Academy of Sciences of the USSR. Andrey Petrovich became the founder of the Siberian school of information science and the academician of Academy of Sciences of the USSR.

Can seem to modern programmers that today it is possible to achieve success in IT and without education, having learned everything independently. As an example of the fact that any hacker self-educated person can achieve success often consider open source. But at the same time forget that the basis of big open source of projects is usually created by professionals with the corresponding education. So, the creator of Linux Linus Torvalds was disaccustomed at the University of Helsinki eight years, having received master degree in computer science. Its final project implemented within the scientific NODES group was called "Linux: A Portable Operating System".

By the way, theoretical knowledge is necessary not only to developers of the complex technical systems, but also creators of applied user applications. The design of applications is not just art, and discipline with deep scientific contents. Such humanitarian fields as semiotics, ontology, human-computer interfaces, psychology, anthropology also are joined to computer science.

Think, for example, why the Chinese website of the producer of the Wrigley chewing gum gives on the homepage references to projects in the field of social responsibility, and the British website meets the visitor by the slogan "We Create Simple Pleasures to Decorate Your Day"? Why on the Italian version of the dating site Match.com one photo of young couple, and on the Venezuelan 24 photos, both young people, and people of middle age?

The theoretical articles based on works of the Netherlands anthropologists and psychologists Fons Trompenaars and Geert Hofstede help to answer these questions. Their models developed in the 70-80th years in the course of studying of cross-cultural communications formed the basis of modern researches of cultural factors in web-design.

It is possible to get acquainted with this analysis in a note^[4]relying on the following scientific publications in more detail:

• Cyr, D. (2008). Modeling Web Site Design Across Cultures: Relationships to Trust, Satisfaction, and E-Loyalty.Journal of Management Information Systems.
• Eristi, S. (2009). Cultural Factors in Web Design. Journal of Theoretical and Applied Information Technology.

And here some more articles illustrating a variety of scientific knowledge of software development:

• Social applications as information source in a workplace^[5]
• Crowdsourcing approach to simplification of the interface of corporate applications^[6]
• Testing approach of graphical interfaces on the basis of the Bayes^[7]

The value of innovations in the field of IT

The few are able even to estimate innovations in the field of IT, not to mention their creation. New chemical process, more effective electric motor, the bioactive molecule killing a dangerous virus — all these inventions form the basis of the new companies and nobody doubts their value. It is worth translating a view of IT startups and the idea that founders understand nothing becomes immutable wisdom and should completely change two or three times the concept of the project.

The biotechnology startup passes a set of stages when the initial idea is confirmed in the development (or not). Each stage — from the idea, computer model, an experiment on yeast and to the 3rd stage of clinical trials is the growth of value of firm and the growing conviction of the market in brilliant perspectives of the project.

Computer and Internet startups still just parade millions of users even if income from each of them (ARPU) is less than a dollar a year. But IT-Innovations can be checked and tested experimentally too. There is a concept of minimum viable product (minimum viable product, MVP), however with its help it is impossible to design difficult, complete solution as the complete solution does not work if only one its part is made. However for some reason methods for similar testing, including sets of standard data and tasks for experiments, remain uninvolved. Scientists in the field of computer science and also their colleagues from related subjects, such as information search, management of databases, library science and information science, human-computer interaction, science about the organization, sociology, cognitive psychology and artificial intelligence, use them in the researches, but developers seldom put even ready results of researches into practice.

GUI of the future is a Terra incognito

Unfortunately, IT professional often ignore scientific results, and not only in Russia, but also abroad.

At a view of forecasts and road maps via interfaces the amusing fact is detected: the subject of graphical interfaces as though is forgotten. In these documents it is possible to meet references of a broad spectrum of gadgets and fashionable words, but nothing is told about what picture will be displayed as a result on screens.

Let's consider for an example: the forecast of Microsoft for 2020^[8] and the Singapore road map on development of IT^[9]. Authors of both documents argue on Internet of Things, haptic interfaces, smart robots and smart clothes, abundance of sensors and the voice interface – and words about how the text, pictures and other virtual objects as if in the future users will be deprived of sight will be displayed (at least directly in an eye). There are, of course, voice assistants, but at the moment they represent the market for specific gadgets, and there are no data that they allow to perform work quicker, more effectively and more conveniently.

At the same time the scientific literature on the subject computer science written in the second half of XX - the beginning of the 21st centuries, contains a set of the revolutionary ideas which potential is almost not used. It is visual language and the structured Robert Horn's letter, the CoDIAK model of Douglas Engelbart, the idea of failure from a modality, the differing interoperability layers with a system, failure from application-centric of interfaces, an object model, a single system of addressing and data storage, the OHS model, the scalable interface of Jef Raskin.

There is a wish to wish to present and future startups not to neglect the deep theoretical preparation which is saved up by base of scientific knowledge in the field of computer science and that range of brilliant ideas which, thanks to the Internet, are available to any who looks for them. Only use of the similar ideas will allow to create really the innovation and much more effective information systems. And only high-quality training of computer science will allow to bring up the professionals capable of it.

Notes