Department of Computer Science, COMSATS University Islamabad, Islamabad Campus, Pakistan.
PLoS One. 2019 Jul 10;14(7):e0219216. doi: 10.1371/journal.pone.0219216. eCollection 2019.
Requirement for formation of collaborations has been on increase for the software development industry, especially for smaller to medium sized firms, due to rapid technological advancements, requirements for diversified skills, ever enhancing demands for innovation and fierce competition. Collaborative product development in an alliance enables the firms to benefit from each other's diversified skills and the experience as a result of which they can develop products more rapidly and of better quality as well resulting in a higher payoff. Also, the development costs decrease. However, to avoid undesired results, selection of an appropriate partner firm for collaboration is of utmost importance keeping in view the objectives of alliance formation of both the strategic partners. One-way partner selection techniques available in the literature are impractical as they enable a firm to rank potential partners only from its own perspective while ignoring their objectives. This problem is addressed by the two-way partner selection techniques, however, they either ignore the payoff distribution criteria or the proposed criteria is unfair. More importantly, existing techniques consider that firm collaborate only with the objective to enhance their financial payoff which might not always be the case. The fact that collaborating firms may have one but different objectives for collaboration, or, each may have multiple objectives is largely neglected. To address the scenarios in which firms may collaborate due to multiple and possibly different objectives, this work proposes a bi-objective game-theoretic model that enables a firm to select an appropriate partner based on the individual preferences of both on the following two objectives: 1) learning and 2) financial revenue. Moreover, this model calculates the pay-off that each firm should get whether only monetary, only in the form of learning or both. The calculation of payoff share is based on the following parameters: 1) individual goals of collaboration of partner selecting firms on the said two objectives, 2) their level of cost contribution, 3) cooperation ratio and 4) knowledge investment difference. Comprehensive analysis of various scenarios is done for the proposed Nash Bargaining payoff distribution model to find the optimum strategy of collaborating firms for each scenario.
由于快速的技术进步、对多元化技能的需求、不断提高的创新需求以及激烈的竞争，软件开发行业，尤其是中小型企业对合作的需求不断增加。联盟中的协作产品开发使企业能够从彼此的多元化技能和经验中受益，从而能够更快地开发出质量更好的产品，并获得更高的回报。此外，开发成本也降低了。然而，为了避免不理想的结果，选择一个合适的合作伙伴进行合作至关重要，同时要考虑到两个战略合作伙伴的联盟形成目标。文献中提供的单向合作伙伴选择技术是不切实际的，因为它们只能使企业从自身角度对潜在合作伙伴进行排名，而忽略了它们的目标。这个问题可以通过双向合作伙伴选择技术来解决，但是，它们要么忽略了收益分配标准，要么提出的标准不公平。更重要的是，现有的技术认为企业合作的目的仅仅是为了提高他们的财务回报，而这并不总是正确的。事实上，合作企业可能有一个但不同的合作目标，或者每个企业可能有多个目标，这些都在很大程度上被忽视了。为了解决企业可能由于多个且可能不同的目标而合作的情况，本工作提出了一个双目标博弈论模型，该模型使企业能够根据以下两个目标：1）学习和 2）财务收益，基于双方的个人偏好选择合适的合作伙伴。此外，该模型还计算了每个企业应该获得的收益，无论是仅仅是货币形式的，仅仅是学习形式的还是两者兼而有之。收益份额的计算基于以下参数：1）合作伙伴选择企业在上述两个目标上的合作个人目标，2）他们的成本贡献水平，3）合作比例和 4）知识投资差异。对所提出的纳什讨价还价收益分配模型的各种情况进行了综合分析，以找到每个场景下合作企业的最优策略。
