Seeds
Social media pulls us closer.
BUT do we always need it?
No data to support your niche topic?
OVERVIEW
When it comes to data analysis and pre-research, we often face the problem of lack of data sources and low quality of surveys. This Q&A and content sharing app incorporates customized data collection/sharing/visualization, using precise recommendations to refer surveys to users who are interested to improve motivation and the quality of responses. The application is an open source site for data, where users can download data on interested topics and do further analysis. Furthermore, the application introduces the Q&AA (anonymous answer) model. Using anonymous answer keyword extraction to quantify comments in order to avoid cyber violence and negative communication in the community.
Kickoff
Early Insights
-
Many valuable topics are lacking data and attention, then it is often necessary to collect data by oneself. And this somehow discourages people to continue those worthwhile topics.
-
For those analyses that require self-collection of data, it is often difficult to find suitable participants to respond.
-
When sharing words on the internet, more and more people receive negative messages harassing them.
OPEN SOURCE
DATA COLLECTION
Q&AA
DATA VISUALIZATION
LINK TO PARTICIPANTS
CONTENT SHARING
LOW NETWORKING
CUSTOMIZATION
COMMENT KEYWORD EXTRACTION
How might it help?
From a business perspective:
-
Help the media (e.g., the news industry) to gather citizens' perceptions and somewhat alternative to interviews.
-
Helping them to generate data visualizations more easily.
-
It can help some small business to alleviate their dilemma of lacking data and better understand the needs of consumers.
-
News media or social influencer can publish articles on it to increase readership.
From a personal perspective:
-
Custom data collection for use in data analysis.
-
Connect to participants who are interested in the topic and therefore return quality responses.
-
Easy to generate visualizations.
-
Away from negative words.
-
Users can have access to articles driven by data on topics of their interest.
Competitor Analysis
Key Insights from Competitor Analysis
There is currently no application on the market that addresses exactly the same problem as Seeds.
-
SurveyMonkey and google form as data collection platforms, the amount of data that may be collected is uncertain as there is no network of questioners and responders. Therefore, they are more suitable for use as internal information collection.
-
Kaggle as an open data source application does not publicly collect data. Besides, it cannot generate data visualization. So that it does not allow you to understand the distribution of data in a short time.
-
Reddit and twitter as social media have a certain polling feature, but you are not able to have direct access to the data and only results can be seen, no further analysis can be done. Also, the polling function they have is relatively simple. At the same time, due to the strong social nature of the applications, there is no anonymity involved.
-
However, Seeds is more focused on data collection as well as content sharing with less emphasis on social networking. That gives me opportunity to add anonymous answers(comments) feature.
USER RESEARCH
1.
Negative attitude of participants.
“The people who fill out the questionnaire are often not self-motivated and so are not very cooperative.”
“Normally, our surveys are sent to people we know to fill out. This is because we can't find more people to fill out the questionnaire, and this way we can guarantee the quality of the responses to a certain extent.”
2.
The participants were chosen from a narrow group.
“When available, we upload our respective surveys to the school's discord to fill in with each other. However, However, this also does not guarantee the quantity and quality of the response.”
“Why not consider placing it on a larger platform?”
“There is no such a platform.”
3.
Lack of (open) data source.
“As long as it is public data are considered easy to find, most difficulties happen when there is no data available.”
“Sometimes a topic comes to mind that I think is good, but once research, there is no data to support it and I feel so discouraged.”
4.
Offenses from strangers can hardly be avoided.
“I try to avoid commenting on others, as I find that mere disagreement can lead to offensive words.”
“I wondered why a lot of apps didn't have anonymity, and then I realized that it might be a protection for me. At least it gives me the opportunity to block him/her.”
DEEPER INSIGHTS
Before I could jump into the designing, it was important define the main pain points behind those interview complaints.
WHY negative attitude?
-
Not feeling involved
-
Users rarely get an update after filling out a survey.
-
This may be because usually participants do not receive the conclusion after filling out the survey. We can recommand the articles on similar topics after the user fills in the survey.
-
And when a user produces content based on the survey, users who have filled out the survey will also receive recommendations for articles. A positive cycle has been established to increase the initiative of users.
-
-
-
Wrong target participants
-
Users are not interested in the topic and thus do not motivated to fill out the survey.
-
By analyzing the user's articles, bookmarks and survey preferences to refer him to surveys on topics that they may be interested in, to drive the quality of survey responses.
-
-
WHY offense can hardly be avoided?
-
Clearly targeted
-
Most social media apps nowadays focus on developing networking features in the hope of strengthening the connection between users, and therefore rarely develop anonymity.
-
We introduced a purely anonymous answering (commenting) approach. The answer section will only show the keywords of answer. We analyze and compare the growth curve of keywords in real time. And keywords that have been mentioned more often are shown on top.
-
-
-
Easy to communicate
-
Direct messages, reply to comments... etc. There are many different ways an unknown user could do to build a connection. Likewise, this is a potential breeding ground for cyberbullying.
-
Users can only see the extracted keywords in the comments section and cannot comment on them again.
-
-
Persona
Mind Map
Sketch
Low-Fidelity Prototypes
Answer questions
Answer
Homepage
Browse
Create a data collection
Collect
Download
Dataset Download
Create
Create a data visualization
Post
Post an article
Insights from usability testing
-
The data download interface is too complex and users have to spend some time to understand it.
-
We can indicate the completion of the data collection to make it easier for users to find the right data set.
-
When the user has only one open-ended question, the involvement of a survey is not necessarily required.
High-Fidelity Prototypes
HOMEPAGE
Users can browse our recommended articles or data collection posts that may be of interest to them, or switch to following mode to see only the topics and the bloggers they follow. In addition, they are also able to switch to 'articles/collections only' mode.
In the preview, the user is informed of the progress of a data collection. At the bottom right corner of each data collection post the number of responses is shown. These help users to select a suitable dataset.
COLLECTION/QUESTION PAGE
Once the data collection details page is opened, the user can participate in the survey. The progress bar above the questions will show the degree of completion, and optional questions will be marked. If an article using this dataset is published in the future, all users who have participated in this data collection will be notified. At the same time, any user can bookmark this data collection post and will be notified when the data collection is complete.
SEARCH PAGE
In the search interface, the application will automatically recommend the topics that users often interest in. In addition, it will also recommend trending topics. The trending topics will show the growth curve of the past 72 hours.
When a user opens a collection post without a survey, adding a comment/answer is considered a participation. This applies to a single open-ended question. The algorithm extracts the keywords based on the user's answers and displays them in the answer field below, and the more frequently mentioned keywords are displayed above, giving them a real-time view of the growth curve for each keyword and the number of times it was mentioned. At the same time, users' full comments are saved and, if they want to access them, they can click on the export data above. Double-clicking on a keyword indicates an agreement, which is also counted as a response in the data collection.
COLLECTION/QUESTION PAGE
WE DO NOT ALLOW USERS TO CHOOSE WHICH RESPONSES TO DOWNLOAD. While they are able to choose the specific attributes to include in his/her own dataset, as well as remove all responses with blank entries.
This is because we want to avoid the cherry-picking as much as possible. Cherry-picking fallacy is the act of capturing data to achieve a desired result. Individual cases or data appear to confirm a particular position while ignoring a significant portion of related and similar cases or data that may contradict that position. We do not support cherry-picking strategy, and therefore user can only download dataset from the first response but not choose in the middle.
WORKSHOP
Once users find the ideal data collection, they can download and save the dataset. Before downloading, they can perform simple filters, such as selecting the attributes, the number of responses they want, and whether to remove responses containing blank entries. While filtering, the user can preview the current dataset in order to get an overview of the data (number, attributes included). The full dataset can be viewed in the workshop.
When everything is ready, then it is ready to be exported. At this point the user can choose to save the dataset to the workshop within seeds, or export it as an exe, zip, etc.
Workshop stores users' datasets and data visualizations, and is the entry to create new data visualizations.
PUBLISH A DATA COLLECTION
Step 1: add title and description
Step 2: create a survey (optional)
When a user decides to publish a data collection, he/she can either choose to publish a (short answer) question or a survey which may contains multiple questions. A data collection usually contains a question title and a description (optional), and if the user chooses not to add a survey, he/she will be redirected to the third step.
If the user chooses to add a survey, then he/she goes to the survey edit page. Here he/she can add questions and options, select the appropriate question type for each question, and determine if it is mandatory. When all questions are finished editing, he/she can go to the next step.
Step 3: Set the time and # of responses
Here users can choose their start and end time for data collection, and the number of responses needed. Users can also choose to have the option of unlimited. They can also be notified when they get the desired number of responses in case of follow-up analysis.
CREATE A DATA VISUALIZATION
After the user selects the dataset, seeds will display the attributes of the dataset accordingly. User can select multiple attributes and the seeds will generate the appropriate data visualizations according to the number and nature of the attributes. At this point, the user can select the desired data visualization, add a title and save it. Of course, the visualization can be further adjusted, such as color adjustment, transpose, add indicator, etc.