MoonFish FAQs

How can I access MoonFish?

MoonFish can be accessed through your BigFish account by clicking on the “+” symbol in the upper left corner of the folder view and selecting “New Taxonomy”. After entering in the taxonomy settings and clicking on “Validate”, you will be redirected to the segment creation menu where you can start building your taxonomy. 

Why aren’t my seedwords coming back with recommendations?

Recommendations are based on how often a word has been seen or mentioned on the internet. If a word doesn’t have enough occurrences, then it will not be possible to generate recommendations for it. 

How can I add a new segment to my taxonomy? 

To add a new segment, click on your taxonomy and select “Add Segment” in the segment breakdown menu. You will then be redirected to the segment creation screen where you will be able to create a new segment. 

Which countries and languages do you cover?

Here is list of the combination of (language, country) handled currently by MoonFish:

• USA, English

• USA, Spanish

• UK, English

• France, French

• Spain, Spanish

• Italy, Italian

• Russia, Russian

• Netherlands, Dutch

• Germany, German

• Portugal, Portuguese

• Brazil, Portuguese

How accurate is the volume estimation?

The precision is higher than 95 %.

What is the freshness of the volume?

The Volume is refreshed once per week.

Is there a limit of words when creating a segment?

You can add a maximum of 100 words to your segment.

How are the recommendations generated?

Recommendations appear ordered by similarity from the seedword. In other words, recommendations are generated according to semantical proximity (similar context).

Example : “Tokyo” and “Paris” have been seen in a similar context as they are both capitals.

To go further in understanding, it is “Word2Vec” which is used. In simple terms, it is a machine learning model that is trained to replicate the linguistic context of words.

I’m trying to create a segment via Keyword List upload. Why isn’t my file uploading?

Please ensure that you are only uploading CSV files in UTF-8 format.

How are the top 100 words computed when creating a segment via Keyword List upload?

The top 100 words are computed based on their similarity to the selected Favorite from the list of matched words. They will appear in descending similarity to the Favorite.

When creating a Moonfish segment, do the seedwords you add have a different weight than the reccomendations? Does the audience estimate depend on these weights? How can we explain this difference and the weight?

The audience estimate isn't linked to the weights of your words, either for seedwords or recommended words. Seedwords do have a different weight by default than your recommended words because they're set as Favorites by default (so with a weight of 1, the maximum), but you can change this by removing their Favorite status if you want.

Reccomendations appear in descending similarity to the seedwords and the color of the bar depends on specificity. What about the length of the bar? Does it depend on the weight of a word in the segment?

Yes it does, the length of the bar represents the weight of the words compared to your Favorite(s) (basically how semantically close they are from your choice of Favorite(s)).

When you click on "Visualize" in order to have a graphic display of a taxonomy, does the dimension of the bubbles depend on the weight of a word?

Yes, that's why you should be able to see that your choices of Favorites will be the larger "bubbles" as they have the heaviest weight.