Periodic Reporting for period 1 - IA-M-SHOP (Intelligent Algorithm for M-Shop)
Reporting period: 2017-09-01 to 2018-11-30
The e-commerce market reached $1,5 Trillion by the end of 2014 at worldwide level. On the other hand, more and more people are just joining the Internet through their Smartphones. So the current e-commerce services will turn into fully m-commerce services in a few years.
For this reason, EUROB has already developed M-SHOP (http://m-shop.io/) an innovative m-commerce platform. However, currently it lacks the integration of an intelligent algorithm for providing personalized offers, which could be key for differentiating the service from new competitors.
In fact, there is no service available in the market specifically designed for Smartphone which already integrates real time personalized offers to their customers. In order to get a unique value proposition respect to competitors, EUROB has made some attempts to develop an intelligent algorithm based on the geo-location of the users, personal profile and preferences, or the purchases history.
In IA-M-Shop project a more in-depth and scientific approach is tackled and with the support of EC and an external expert associated previous works are taken to the next step.
In particular the following objectives have been tackled:
- Design and develop an intelligent algorithm based on the user’s location (GPS coordinates), personal profile, preferences, purchase history, browse history, IoT sensor data (temperature, probability of rain, etc.) in order to provide the best purchase recommendations.
- Identify additional income sources based on the intelligent algorithm. Design a business plan for each of them. Investigate on potential adaptions of the intelligent algorithm to EUROB’s future or already developed products.
- Identify the state-of-the-art of similar algorithms and establish the differences in the technological aspects of our solution respect to competitors. Deeply analyse actual and future competitors. Elaborate the best possible IPR strategy for IA-MSHOP.
Some of the key functionalities addressed in IA-SHOP are:
- Urban Integration: the algorithm will include environmental information from IoT sensors;
- Social Shopping: the algorithm will take into account purchases from acquittances, friends and family
- Privacy and Encryption: the service architecture must take into account GDPR requirements and privacy and security by design.
- Personalized Offers: Offers should be personalized and individual per user, tailor made.
For this reason, EUROB has already developed M-SHOP (http://m-shop.io/) an innovative m-commerce platform. However, currently it lacks the integration of an intelligent algorithm for providing personalized offers, which could be key for differentiating the service from new competitors.
In fact, there is no service available in the market specifically designed for Smartphone which already integrates real time personalized offers to their customers. In order to get a unique value proposition respect to competitors, EUROB has made some attempts to develop an intelligent algorithm based on the geo-location of the users, personal profile and preferences, or the purchases history.
In IA-M-Shop project a more in-depth and scientific approach is tackled and with the support of EC and an external expert associated previous works are taken to the next step.
In particular the following objectives have been tackled:
- Design and develop an intelligent algorithm based on the user’s location (GPS coordinates), personal profile, preferences, purchase history, browse history, IoT sensor data (temperature, probability of rain, etc.) in order to provide the best purchase recommendations.
- Identify additional income sources based on the intelligent algorithm. Design a business plan for each of them. Investigate on potential adaptions of the intelligent algorithm to EUROB’s future or already developed products.
- Identify the state-of-the-art of similar algorithms and establish the differences in the technological aspects of our solution respect to competitors. Deeply analyse actual and future competitors. Elaborate the best possible IPR strategy for IA-MSHOP.
Some of the key functionalities addressed in IA-SHOP are:
- Urban Integration: the algorithm will include environmental information from IoT sensors;
- Social Shopping: the algorithm will take into account purchases from acquittances, friends and family
- Privacy and Encryption: the service architecture must take into account GDPR requirements and privacy and security by design.
- Personalized Offers: Offers should be personalized and individual per user, tailor made.
In IA-M-SHOP project a first design and development of an intelligent recommendation algorithm has been proposed based on the latest SoA and technical results. It has been customized for the specific use case of M-Shop and Incity Together products of the company so it can be applied based on the actual results and SW of the company and it features an intelligent scalable design.
The State-of-the-Art of existing algorithms related to e-commerce and m-commerce was analysed. Concluding that there are many similarities between E-commerce and m-commerce but also differential characteristics. Some examples of m-commerce platforms (potential competitors) were identified:
1. Piccadilly Records
2. The Gadget Flow
3. Net-A-Porter
4. Threadless
5. Curry’s
In order to target adequately the right recommendations to overcome the known problem of ‘cold starting’ for recommendations algorithms, and in absence of previous relevant data from the user, a set of statistical assumptions were made as baseline status for the intelligent algorithm start. For this the sociodemographic profiling of Spanish people who have made online purchases by their gender, age, level of studies, social layer and size of the place they live in was mostly obtained from INE’s database and reports.
Related to the algorithm parameters selection, after a deep study, it was established the following major categories for the parameters being considered in the calculations:
- Event Tickets
- Retail articles
- Travel offers
The works reached the level of the architecture design of the algorithm and modules to be developed.
The State-of-the-Art of existing algorithms related to e-commerce and m-commerce was analysed. Concluding that there are many similarities between E-commerce and m-commerce but also differential characteristics. Some examples of m-commerce platforms (potential competitors) were identified:
1. Piccadilly Records
2. The Gadget Flow
3. Net-A-Porter
4. Threadless
5. Curry’s
In order to target adequately the right recommendations to overcome the known problem of ‘cold starting’ for recommendations algorithms, and in absence of previous relevant data from the user, a set of statistical assumptions were made as baseline status for the intelligent algorithm start. For this the sociodemographic profiling of Spanish people who have made online purchases by their gender, age, level of studies, social layer and size of the place they live in was mostly obtained from INE’s database and reports.
Related to the algorithm parameters selection, after a deep study, it was established the following major categories for the parameters being considered in the calculations:
- Event Tickets
- Retail articles
- Travel offers
The works reached the level of the architecture design of the algorithm and modules to be developed.
It has been pending to port to code and validate in the company software the designed algorithm, due to the associate not being able to complete the full project.
The company is completing the works based on internal self effort
The company is completing the works based on internal self effort