Numerous organizational meetings have taken place during the formation of OAITI. Often times, the length of these meetings would dictate that we have some snacks to go along with them. These would often be some type of food bar, such as a LARABAR:
Many times, we’d informally rate these “The taste is great, but I’m not a big fan of the texture!”, “It has an interesting aftertaste”, “The branding is a little boring”. As more meetings took place, we realized that it might be worth our time to formally collect ratings data on each bar that we try as a group. As the months went on, we collected enough data to do some rudimentary analysis of our preferences. The data in its complete raw form is given below:
| Food Bar | Taster | Texture | Natural | Taste | BarAppearance | Branding | AfterTaste-Potency[-10+10] | Smell[-10+10] | Overall | Comments | Date |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Larabar - Ginger Bread | Lawrence | 6.5 | 8 | 6.75 | 7 | 42760 | |||||
| Larabar - Ginger Bread | Eric | 7 | 9 | 6 | 7 | 42760 | |||||
| Larabar - Ginger Bread | Nadia | 8 | 8 | 6 | 7 | 42760 | |||||
| Health Warrior | Eric | 4 | 8 | 6 | 3 | 9 | 1 | 6 | 42765 | ||
| Health Warrior | Lawrence | 5.5 | 9 | 4.5 | 3.5 | 6.5 | 7 | 5 | 42765 | ||
| Health Warrior | Nadia | 7 | 9 | 8 | 6 | 7 | 9 | 8 | 42765 | ||
| Health Warrior | Pete | 3 | 8 | 5 | 3 | 7 | -3 | 4 | weird | 42765 | |
| Power Crunch | Lawrence | 9.5 | 2 | 7.5 | 6 | 5.5 | -0.5 | 6.5 | 42800 | ||
| Power Crunch | Eric | 9 | 3 | 8 | 4 | 6 | 1 | 7 | goes well with cactus water | 42800 | |
| Power Crunch | Nadia | 9 | 2.5 | 9 | 4 | 4 | -5 | 5 | 42800 | ||
| Muscle Milk | Eric | 7 | 2 | 4 | 7 | 8 | -3 | 4 | 42807 | ||
| Muscle Milk | Lawrence | 3 | 5.5 | 3.5 | 7 | 8.5 | 1 | 3.5 | 42807 | ||
| Muscle Milk | Nadia | 7 | 3 | 3 | 6 | 7 | 4 | 3 | 42807 | ||
| Nutrition-Zone Perfect | Nadia | 9 | 8 | 9 | 7 | 8 | 3 | 5 | 9 | crunchy | 42810 |
| Nutrition-Zone Perfect | Eric | 8 | 7 | 9 | 7 | 9 | 5 | 8 | 9 | possibly favourite | 42810 |
| Nutrition-Zone Perfect | Lawrence | 1.5 | 5 | 8 | 6 | 9 | 2 | 9.5 | 7.5 | 42810 | |
| Kind | Eric | 4 | 8.5 | 8 | 3 | 6 | 0 | 0 | 8 | 42814 | |
| Kind | Nadia | 7 | 9 | 8 | 8 | 6 | 5 | 7 | 9 | crunchy! | 42814 |
| Kind | Lawrence | 5.5 | 8.5 | 5.5 | 3 | 6.5 | 8 | 0 | 6 | 42814 | |
| Cliff Bar | Nadia | 6.5 | 7 | 7 | 4 | 7 | -1 | 7 | 6.5 | natural but protein powder taste | 42817 |
| Cliff Bar | Eric | 4 | 3 | 7 | 2.5 | 5 | 0 | 8 | 6 | 42817 | |
| Cliff Bar | Lawrence | 6.5 | 4.5 | 6.5 | 4 | 5 | -0.5 | 7 | 7 | 42817 | |
| Combat Birthday Cake | Eric | 3 | 2 | 2 | 4 | 3.5 | -5 | -6 | 2 | Protien taste! | 42824 |
| Combat Birthday Cake | Nadia | 2 | 0 | 3 | 7 | 5 | -3 | -4 | 1 | Odd brand | 42824 |
| Combat Birthday Cake | Lawrence | 3.5 | 3 | 2.5 | 6.5 | 3 | -3 | -2 | 2 | 42824 | |
| SMART- CookieDough | Nadia | 5 | 3 | 4 | 7 | 6 | -1 | 3 | 4 | 42828 | |
| SMART- CookieDough | Eric | 7 | 8.5 | 8 | 7.5 | 9.5 | 2 | 6 | 8 | 42828 | |
| SMART- CookieDough | Lawrence | 6.5 | 3.5 | 4.5 | 8.5 | 7.5 | 1.5 | 6.5 | 5 | 42828 | |
| SMART -Blueberry | Nadia | 5 | 5 | 6 | 7 | 6 | 2 | -1 | 6 | 42831 | |
| SMART -Blueberry | Eric | 6 | 8 | 6.5 | 7 | 9 | 0 | -4 | 6 | 42831 | |
| SMART -Blueberry | Lawrence | 6.5 | 7 | 7.5 | 7.5 | 7.5 | 0.5 | -2 | 7.5 | 42831 | |
| Bounce | Lawrence | 5 | 3.5 | 5 | 5 | 9 | 5 | 5.5 | 8.5 | an experience | 42835 |
| Bounce | Eric | 7 | 7 | 8 | 6 | 9 | 6 | 3 | 8.5 | 42835 | |
| Bounce | Nadia | 8 | 8 | 8 | 9 | 9 | -1 | -1 | 8 | unusual | 42835 |
| Health Warrior | Nadia | 7 | 7 | 7 | 5 | 6 | 1 | -2 | 7 | 42838 | |
| Health Warrior | Lawrence | 7 | 9 | 6.5 | 3 | 6.5 | 1 | -6 | 7.5 | 42838 | |
| Health Warrior | Eric | 6 | 8 | 6 | 3 | 7 | 0 | -4 | 6.5 | 42838 | |
| Sahale Snacks - Orange | Nadia | 6 | 7.5 | 7.5 | 7 | 8 | 5 | 1 | 7 | 42845 | |
| Sahale Snacks - Orange | Lawrence | 6.5 | 7.5 | 4 | 7.5 | 6.5 | 1 | -1.5 | 5 | 42845 | |
| Sahale Snacks - Orange | Eric | 5 | 9 | 8 | 6 | 7 | 4 | 0 | 7.5 | 42845 | |
| Sahale Snacks - Banana Pecan | Nadia | 7 | 6 | 7.5 | 6.5 | 7 | -1 | 3 | 7 | 42849 | |
| Sahale Snacks - Banana Pecan | Lawrence | 7 | 7.5 | 6.5 | 8 | 7 | 2 | 1 | 7 | 42849 | |
| Sahale Snacks - Banana Pecan | Eric | 7.5 | 9 | 8.5 | 6.5 | 8.5 | 6 | 2 | 8.5 | 42849 | |
| Good Greens | Nadia | 7 | 6 | 8 | 6 | 7 | 1 | -4 | 7.5 | 42852 | |
| Good Greens | Lawrence | 5 | 6.5 | 4 | 5 | 7.5 | 0.5 | -1 | 5 | 42852 | |
| Good Greens | Eric | 5 | 8 | 4.5 | 2 | 9 | -3 | -6 | 4 | 42852 | |
| Vans- PBJ | Nadia | 7 | 7 | 8.5 | 7 | 8 | 6 | 2 | 8.5 | 42870 | |
| Vans- PBJ | Lawrence | 8.5 | 6.5 | 9 | 7.5 | 8 | -1 | 3.5 | 8.5 | 42870 | |
| Vans- PBJ | Hieu | 4 | 5 | 5 | 7 | 8 | 3 | 0 | 7 | 42870 | |
| Vans- PBJ | Eric | 7 | 7 | 8 | 7.5 | 9 | 2 | 5 | 8 | 42871 | |
| Larabar - Key Lime Pie | Eric | 8 | 8 | 6.5 | 4 | 4 | 2 | -3 | 7 | 42871 | |
| Larabar - Key Lime Pie | Nadia | 7 | 7 | 7 | 6 | 4 | 0.5 | 2 | 7.5 | 42871 | |
| Larabar - Key Lime Pie | Lawrence | 7 | 7.5 | 7 | 5.5 | 3.5 | 0 | 0.5 | 7 | 42871 | |
| OhYeah - One | Eric | 5 | 3.5 | 4.5 | 7.5 | 8.5 | -4 | 5.5 | 4 | 42873 | |
| OhYeah - One | Nadia | 3 | 3 | 6 | 7 | 8 | -2 | 7 | 5.5 | 42873 | |
| OhYeah - One | Lawrence | 5 | 2 | 2 | 9 | 3.5 | -2 | 2.5 | 4 | 42873 | |
| GoMacro - Protein Purity | Nadia | 7 | 7 | 5 | 4 | 8 | -1 | -3 | 4 | 42874 | |
| GoMacro - Protein Purity | Lawrence | 6.5 | 8 | 5 | 4 | 9 | -4 | -1 | 3.5 | 42874 | |
| GoMacro - Protein Purity | Hieu | 4 | 8 | 5 | 0 | 8 | -10 | -5 | 3 | 42874 | |
| GoMacro - Protein Purity | Eric | 6 | 9 | 5 | 3.5 | 9 | -1 | -1 | 5.5 | 42874 | |
| GoMacro - Protein Pleasure | Nadia | 8 | 8 | 7 | 8 | 8 | 1 | 2 | 7 | 42874 | |
| GoMacro - Protein Pleasure | Lawrence | 5 | 5.5 | 3.5 | 7.5 | 9 | -2 | -0.5 | 4.5 | 42874 | |
| GoMacro - Protein Pleasure | Hieu | 6 | 7 | 6 | 0 | 8 | 4 | -5 | 7 | 42874 | |
| GoMacro - Protein Pleasure | Eric | 7 | 8 | 6 | 4.5 | 9 | 3 | 1 | 7.5 | 42874 | |
| GoMacro - Balanced Goodness | Nadia | 6 | 7 | 8 | 9 | 8 | 6 | 5 | 8 | 42874 | |
| GoMacro - Balanced Goodness | Lawrence | 7.5 | 8.5 | 9 | 7.5 | 8.5 | -1 | 3 | 9 | 42874 | |
| GoMacro - Balanced Goodness | Hieu | 7 | 9 | 9 | 5 | 8 | 9 | -5 | 8.5 | 42874 | |
| GoMacro - Balanced Goodness | Eric | 5.5 | 9 | 6.5 | 5.5 | 9 | 3 | 2 | 7 | 42874 | |
| Health Warrior-Coconut | Lawrence | 7.5 | 8 | 3.5 | 5.5 | 7.5 | -0.5 | -3 | 4.5 | 42891 | |
| Health Warrior-Coconut | Eric | 8 | 8.5 | 8.5 | 6 | 7 | 2 | 3 | 8.5 | 42891 | |
| Health Warrior-Coconut | Nadia | 8 | 8 | 7 | 6 | 7 | 1 | 1 | 7.5 | 42891 | |
| Sunny Days- Strawberry | Hieu | 4 | 7 | 7 | 4 | 5 | 5 | 10 | 8 | 42905 | |
| Sunny Days- Strawberry | Eric | 5 | 4.5 | 6 | 5.5 | 4 | 0 | 8 | 5.5 | 42905 | |
| Sunny Days- Strawberry | Nadia | 4 | 3 | 6 | 7 | 8 | -2 | 8 | 7 | 42905 | |
| Sunny Days- Strawberry | Lawrence | 8 | 8 | 7.5 | 3 | 2.5 | 0.5 | 4 | 7 | 42905 | |
| Kashi-Almond Snickerdoodle | Hieu | 10 | 6 | 9 | 8 | 3 | 6 | 8 | 42909 | ||
| Kashi-Almond Snickerdoodle | Eric | 8.5 | 9 | 9 | 8.5 | 5 | 5 | 8.5 | 42909 | ||
| Kashi-Almond Snickerdoodle | Nadia | 6 | 8 | 7 | 7 | 8 | -2 | 3 | 7 | 42909 | |
| Kashi-Almond Snickerdoodle | Lawrence | 5 | 7.5 | 7 | 7.5 | 5.5 | 1 | 8.5 | 7 | 42909 | |
| Whenever Bars - Oat Blueberry Lemon | Eric | 6 | 8 | 9 | 4.5 | 9 | 2 | 3 | 8.5 | 42912 | |
| Whenever Bars - Oat Blueberry Lemon | Nadia | 4 | 6 | 8 | 6 | 8 | -4 | 2 | 8 | 42912 | |
| Whenever Bars - Oat Blueberry Lemon | Lawrence | 2.5 | 7.5 | 5.5 | 2 | 7.5 | 1 | 3.5 | 6.5 | 42912 |
You’ll immediately note some issues with the data:
- We added some additional attributes (BarAppearance, Branding, AfterTaste, Smell) after we began collecting data, so these fields are missing for the first bar (or in the case of Smell, the first several bars)
- Not every member of the team has rated every bar, so there is missing data. (In addition, we had some guest tasters who have only rated one or two bars total)
- The date field is transformed into an integer – This needs to be converted to be an actual date (as a matter of fact, this integer is the number of days since December 31st, 1899)
Using this data, and the R package recommenderlab, we created a Shiny application which attempts to fill in the missing ratings for particular users, by using their ratings to form recommendations for the bars they didn’t rate. Similar (but far more advanced) recommender systems are used by Netflix, Amazon, and many other online services and retailers in order to recommend products to users. We felt this was a neat prototype of a recommender system partly because of its simplicity, and partly because of the self-collected nature of the data. A screenshot of the application is available below:
The app is available at the following URL: https://oaiti.org/apps/foodbars/ – Feel free to explore the data yourself. A modern web browser, such as the latest version of Google Chrome, Mozilla Firefox, or Microsoft Edge is required. The full source code application and data processing, is available on our GitHub Page.
Leave a Reply
You must be logged in to post a comment.