Consumer insights analysis using word embeddings
In one embodiment, a method includes receiving a request to generate k keywords each of which is semantically related to a particular subject, where the request includes an input n-gram representing the particular subject, accessing a table of word vector relationships, where the table includes a plurality of unique n-grams and their corresponding word vectors, and wherein each of the word vectors represents a semantic context of a corresponding n-gram as a point in a d-dimensional embedding space, looking up, using the table, a first word vector corresponding to the input n-gram, selecting k word vectors closest to the first word vector in the embedding space using the table and based on a similarity metric, identifying, for each of the selected word vectors, a corresponding n-gram by looking up the selected word vector in the table, and sending a response message including the identified n-grams.
1. A method comprising:
by a first computing device in an online social network, receiving, from a second computing device, a request to generate k keywords each of which is semantically related to a particular subject in public perceptions of a particular demographic, wherein the request comprises an input n-gram representing the particular subject, and wherein the request comprises one or more conditions characterizing the particular demographic;
by the first computing device, constructing a corpus of text by collecting text content from content objects created by users of the online social network who satisfy the one or more conditions;
by the first computing device, constructing a table of word vector relationships by training a word embedding mode using the constructed corpus of text as training data, wherein the table comprises a plurality of unique n-grams and their corresponding word vectors, and wherein each of the word vectors represents a semantic context of a corresponding n-gram as a point in a d-dimensional embedding space;
by the first computing device, looking up, using the table, a first word vector corresponding to the input n-gram;
by the first computing device, selecting, using the table and based on a similarity metric, k word vectors closest to the first word vector in the embedding space;
by the first computing device, identifying, for each of the selected word vectors, a corresponding n-gram by looking up the selected word vector in the table; and
by the first computing device, sending, to the second computing device, a response message comprising the identified n-grams.
2. The method of claim 1 , wherein the particular subject is a person, an object, or a concept.
3. The method of claim 1 , wherein the plurality of unique n-grams in the table are selected from the corpus of text.
4. The method of claim 1 , wherein the word embedding model is a word2vec model.
5. The method of claim 1 , wherein the looking up a word vector corresponding to the input n-gram comprises looking up the n-gram in the table.
6. The method of claim 1 , wherein the similarity metric is a cosign similarity, a Euclidean distance, or a Jaccard similarity coefficient.
7. The method of claim 1 , wherein the content objects were created within a pre-determined period of time.
8. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:
receive, from a second computing device, a request to generate k keywords each of which is semantically related to a particular subject, wherein the request comprises an input n-gram representing the particular subject in public perceptions of a particular demographic, and wherein the request comprises one or more conditions characterizing the particular demographic;
construct a corpus of text by collecting text content from content objects created by users of the online social network who satisfy the one or more conditions;
construct a table of word vector relationships by training a word embedding model using the constructed corpus of text as training data, wherein the table comprises a plurality of unique n-grams and their corresponding word vectors, and wherein each of the word vectors represents a semantic context of a corresponding n-gram as a point in a d-dimensional embedding space;
look up, using the table, a first word vector corresponding to the input n-gram;
select, using the table and based on a similarity metric, k word vectors closest to the first word vector in the embedding space;
identify, for each of the selected word vectors, a corresponding n-gram by looking up the selected word vector in the table; and
send, to the second computing device, a response message comprising the identified n-grams.
9. The media of claim 8 , wherein the particular subject is a person, an object, or a concept.
10. The media of claim 8 , wherein the plurality of unique n-grams in the table are selected from the corpus of text.
11. The media of claim 8 , wherein the word embedding model is a word2vec model.
12. The media of claim 8 , wherein the looking up a word vector corresponding to the input n-gram comprises looking up the n-gram in the table.
13. The media of claim 8 , wherein the similarity metric is a cosign similarity, a Euclidean distance, or a Jaccard similarity coefficient.
14. A system comprising:
one or more processors; and
one or more computer-readable non-transitory storage media coupled to one or more of the processors and comprising instructions operable when executed by one or more of the processors to cause the system to:
receive, from a second computing device, a request to generate k keywords each of which is semantically related to a particular subject in public perceptions of a particular demographic, wherein the request comprises an input n-gram representing the particular subject, and wherein the request comprises one or more conditions characterizing the particular demographic;
construct a corpus of text by collecting text content from content objects created by users of the online social network who satisfy the one or more conditions;
construct a table of word vector relationships by training a word embedding model using the constructed corpus of text as training data, wherein the table comprises a plurality of unique n-grams and their corresponding word vectors, and wherein each of the word vectors represents a semantic context of a corresponding n-gram as a point in a d-dimensional embedding space;
look up, using the table, a first word vector corresponding to the input n-gram;
select, using the table and based on a similarity metric, k word vectors closest to the first word vector in the embedding space;
identify, for each of the selected word vectors, a corresponding n-gram by looking up the selected word vector in the table; and
send, to the second computing device, a response message comprising the identified n-grams.
15. The system of claim 14 , wherein the particular subject is a person, an object, or a concept.
16. The system of claim 14 , wherein the plurality of unique n-grams in the table are selected from the corpus of text.
17. The system of claim 14 , wherein the word embedding model is a word2vec model.
18. The system of claim 14 , wherein the looking up a word vector corresponding to the input n-gram comprises looking up the n-gram in the table.
19. The system of claim 14 , wherein the similarity metric is a cosign similarity, a Euclidean distance, or a Jaccard similarity coefficient.