Following the first hackathon, which was well received in May 2015 under the theme of "Issues that are there now," the second artificial intelligence hackathon was held for two days from December 5th to 1th.The theme this time is "food".And, as a guest mentor, we welcomed Mr. Masato Taruishi, CTO of "Retty", which is a hot topic as a real name type gourmet service.Mr. Taruseki not only participated as a mentor, but also provided some of the word-of-mouth and shop information of "Retty", and analyzed this data with FRONTEO's artificial intelligence "KIBIT", which is a highly realistic initiative. About 2 people from 12 teams selected from a large number of applications challenged. After two days of development, services full of ideas were announced one after another, and it was a two-day period with a high sense of fulfillment and accomplishment.
■ "Food" triggers the generation of ideas for understanding and solving various problems.
The theme of this hackathon was "food," which is familiar to everyone, and it was a major feature that all teams were more specific and easy to understand the direction of problem setting and solution.On the other hand, there is a bias in the direction of service due to strong feelings, and in the mentoring on the evening of the first day, there are more problems of data collection, issues as a business model, than Takeda of Taruseki CTO and FRONTEO. Severe points such as whether it is a service that people can use continue, and there are scenes where participants bite their lips ...
■ After a number of presentations full of ideas, the screening announcement is finally over!
Mentoring with Mr. Taruseki and Takeda who are enthusiastic
On the second day, after about 2 hours of ideathon and development, the presentation was attended by two mentors as judges and Foreign Deb Hasshi (Yo Hashimoto), a gourmet entertainer, for 20 minutes for each team. In addition to showing the analysis results in their time, there was a variety of presentations such as skits and introductions in applications on smartphones.The results of the examination and the reasons for the award are as follows.
Foreign Fat Award
Team "Gluttony Mansai" [Details] A second party that bothers the secretary.Based on the store information such as the number of participants and business hours, and the direction of the store you want to go to, we will answer the store that seems to match in 3 seconds. [Reason for award] The reason for the award was that it was very attractive throughout, such as finding a shop for the second party and how to make a deep system. (Mr. Hasshi)
Mr. Hasshi (far left) and the team "Gluttony Mansai"
Team "SOOS" [Details] Create an "analyzer" based on the information written on overseas word-of-mouth sites, machine-translate the information written by Retty, and give advice to foreigners visiting Japan when choosing a store. [Reason for award] It matches Retty's mission of "making the world happy through food," and is very enthusiastic about "creating value not only for people in Japan but for people all over the world." I sympathized. (Mr. Taruseki)
Mr. Taruseki (center) and team "SOOS"
Ken Ninomiya of the team "Super NANIKA" [Details] For stores that want to be diagnosed, such as the number of customers, based on reviews, clarify "elements that are lacking compared to other stores" and give advice to increase the number of customers. Connect to. [Reason for award] In the analysis, I thought about the characteristics of KIBIT.On top of that, we added value to our services from a new perspective and created new indicators. (Takeda)
Takeda (left) and Ken Ninomiya
Team "Diversity" [Reason for award] The point of view was good: service ideas, smooth chat and recommendations. Good results have also been obtained in the analysis of store selection using KIBIT.It also overlapped with what we were trying to achieve using UBIC's artificial intelligence-equipped robot "Kibiro." (Takeda)
Takeda and the team "Diversity" that won the highest award
The service proposed by the team "Diversity", which won the highest award, was to collect people's opinions as texts via bots on chat and propose recommended stores based on the text information.It seems that not only the judges but also the participants were sympathetic to the high degree of perfection that anyone can imagine the usage scene without difficulty, and because it is the time when the year-end party season is just around the corner, "Implemented tomorrow I want you to do it. "This time, the other teams who unfortunately missed the award also tried to solve the practical problems in "food" and "selecting a store".Ideas with high practicality are born, such as presenting shops that can be used on dates and making recommendations by understanding the subtle differences in ramen tastes.Finally, Takeda said that it was a very meaningful hackathon, with the impression that "I think that the participants deepened their understanding of the API compared to the previous hackathon, and I think that more realistic ideas and highly accurate analysis results were obtained." I did. At FRONTEO, we will continue to have practicality like this, and we would like to work together through hackathons to come up with ideas for services that everyone wants.Thank you to everyone who participated and the staff!
After the awards ceremony, everyone took a commemorative photo
FRONTEO's artificial intelligence "KIBIT"
Robot with artificial intelligence KIBIT Kibiro