Using a Kano-like model to facilitate open innovation in requirements engineering
Date
2019-11-01
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Kui viiakse läbi nõuete analüüsi (inglise k Requirements Engineering, lühend RE), siis sageli järjestatakse nõuded nende olulisuse alusel (inglise k requirements prioritization), et saada selgust, milliste välja pakutud nõuetega funktsioon peaks tarkvaral olemas olema, seega sõltub tarkvara analüüsist tarkvara majandusliku väärtuse suurendamisega seotud otsuste tegemine. Tänapäeval arenevad tooted väga kiiresti ning ka nõuete olulisuse alusel järjestamine (inglise k requirements prioritization) on muutunud kiiremaks. Ettevõtted sooviksid saada kasutajatelt kiiret tagasisidet selle kohta, mis peaks olema järgmises mudelis olemas. Üks häid lahendusi sellele on Kano mudel (inglise k Kano model). Kano mudel selgitab välja kasutajate rahulolu ja toodete tunnuste vahelise suhte. See meetod liigitab kasutajate eelistused nende tähtsuse järjekorras, seega toetab see ka nõuete olulisuse järjekorra moodustamist. Aga Kano mudeli rakendamine on kallis ja aeganõudev ning seda ei saa kiiresti korrata. Veelgi enam – see mudel on keeruline väikeste ettevõtete jaoks, sest neil ei tarvitse olla piisavalt rahalisi jm vahendeid, et kasutajatega ühendust võtta ja neid intervjueerida. See omakorda paneb väikesed ettevõtted, eriti just idufirmad, ebavõrdsesse olukorda suurte ettevõtetega.
Et sellele probleemile lahendust leida ja Kano mudeli kasutuselevõttu lihtsamaks ning odavamaks teha, arvame, et Kano mudelit tuleks arendada kahel viisil. Esiteks tuleks kasutada tasuta võrgus leiduvaid kirjalikke andmeid, mida saaks asendada intervjueeritavatelt kogutud vastustega. Teiseks – selleks, et hakkama saada võrgust kogutud kirjalike andmete suure mahuga, ning et kaasa aidata korrapärastele analüüsidele, peaks andmete analüüsimine olema automaatne.
Selle uurimuse eesmärk on välja pakkuda meetodeid, et kasutajate avamusi, mis on võrgus saadavatest vabadest allikatest kogutud, (semi-)automaatselt liigitada, ja seda selleks, et aidata otsustajatel otsustada, millised tarkvara nõuded järgmises mudelis kindlasti olemas peaksid olema. Et seda uurimuse eesmärki saavutada, pakume me välja avatud innovatsiooni nõuete analüüsi (OIRE) meetodi, mille abil saavad tarkvarafirmad parema ülevaate kasutajate vajadustest ja sellest, kuivõrd rahul on nad olemasolevate toodetega.
When Requirements Engineering (RE) is applied, requirements analysis is often used to determine which candidate requirements of a feature should be included in a software release. This plays a crucial role in the decisions made to increase the economic value of software. Nowadays, products evolve fast, and the process of requirements prioritization is becoming shorter as well. Companies benefit from receiving quick feedback from end users about what should be included in subsequent releases. One effective approach supporting requirements prioritization is the Kano model. The Kano model defines the relationship between user satisfaction and product features. It is a method used to classify user preferences according to their importance, and in doing so, supports requirements prioritization. However, implementing the Kano model is costly and time-consuming, and the application of the Kano model cannot be repeated quickly. Moreover, this is even more difficult for small companies because they might not have sufficient funds and resources to contact end users and conduct interviews. This puts small businesses, especially start-ups, at an unfair disadvantage in competing with big companies. To address this problem and make the application of the Kano model simpler, faster, and cheaper, we propose evolving the Kano model in two aspects. First, free online text data should be used to replace responses collected from interviewees. Second, in order to handle the higher amount of data that can be collected from free online text data and in order to facilitate frequent analyses, the data analysis process should be automated. The goal of this research is to propose methods for (semi-)automatically classifying user opinions collected from online open sources (e.g., from online reviews) to help decision-makers decide which software requirements to include in subsequent product versions. To achieve this research goal, we propose the Open Innovation in Requirements Engineering (OIRE) method to help software organizations gain a better understanding of user needs and satisfaction with existing products. A key element of the OIRE method is its Kano-like model. This Kano-like model mimics the traditional Kano model, except that it uses data from online reviews instead of interviews conducted with select focus groups.
When Requirements Engineering (RE) is applied, requirements analysis is often used to determine which candidate requirements of a feature should be included in a software release. This plays a crucial role in the decisions made to increase the economic value of software. Nowadays, products evolve fast, and the process of requirements prioritization is becoming shorter as well. Companies benefit from receiving quick feedback from end users about what should be included in subsequent releases. One effective approach supporting requirements prioritization is the Kano model. The Kano model defines the relationship between user satisfaction and product features. It is a method used to classify user preferences according to their importance, and in doing so, supports requirements prioritization. However, implementing the Kano model is costly and time-consuming, and the application of the Kano model cannot be repeated quickly. Moreover, this is even more difficult for small companies because they might not have sufficient funds and resources to contact end users and conduct interviews. This puts small businesses, especially start-ups, at an unfair disadvantage in competing with big companies. To address this problem and make the application of the Kano model simpler, faster, and cheaper, we propose evolving the Kano model in two aspects. First, free online text data should be used to replace responses collected from interviewees. Second, in order to handle the higher amount of data that can be collected from free online text data and in order to facilitate frequent analyses, the data analysis process should be automated. The goal of this research is to propose methods for (semi-)automatically classifying user opinions collected from online open sources (e.g., from online reviews) to help decision-makers decide which software requirements to include in subsequent product versions. To achieve this research goal, we propose the Open Innovation in Requirements Engineering (OIRE) method to help software organizations gain a better understanding of user needs and satisfaction with existing products. A key element of the OIRE method is its Kano-like model. This Kano-like model mimics the traditional Kano model, except that it uses data from online reviews instead of interviews conducted with select focus groups.
Description
Keywords
software development, requirements engineering, open innovation, feedback, data mining, quality management