twitter-sentiment

Twitter sentiment analysis using Language API connector and iteration syntax

In this sample, you will see how to use Cloud Natural Language API connector and iteration syntax to analyze sentiments of latest tweets of a Twitter handle.

Twitter API

In order to use Twitter API, you need to sign up for a developer account. Once you have the account, you need to create an app and get a bearer token to use in your API calls. This tutorial assumes that you already have a bearer token ready.

Twitter has an API to search for Tweets. Here's an example to get 100 Tweets from @GoogleCloudTech:

BEARER_TOKEN=...
TWITTER_HANDLE=GoogleCloudTech
MAX_RESULTS=100

curl -X GET -H "Authorization: Bearer $BEARER_TOKEN" "https://api.twitter.com/2/tweets/search/recent?query=from:$TWITTER_HANDLE&max_results=$MAX_RESULTS"

Cloud Natural Language API

Natural Language API uses machine learning to reveal the structure and meaning of texts. It has methods such as sentiment analysis, entity analysis, syntactic analysis and more.

In this example, you will use sentiment analysis. Sentiment analysis inspects the given text and identifies the prevailing emotional opinion within the text, especially to determine a writer's attitude as positive, negative, or neutral.

You can see a sample sentiment analysis response here. You will use score of documentSentiment to identify the sentiment of each post. Score ranges between -1.0 (negative) and 1.0 (positive) and corresponds to the overall emotional leaning of the text.

You will also calculate average and minimum sentiment score of all processed tweets.

Enable services

First, enable required services:

gcloud services enable \
  workflows.googleapis.com \
  language.googleapis.com

Define workflow

Create a workflow.yaml to define the workflow.

In the init step, read the bearer token, twitter handle, max results for the Twitter API. We also keep track of some sentiment variables:

main:
  params: [args]
  steps:
    - init:
        assign:
          - bearerToken: ${args.bearerToken}
          - twitterHandle: ${args.twitterHandle}
          - maxResults: ${args.maxResults}
          - totalSentimentScore: 0
          - minSentimentScore: 1
          - minSentimentIndex: -1

In the searchTweets step, fetch Tweets using the Twitter API:

    - searchTweets:
        call: http.get
        args:
          url: ${"https://api.twitter.com/2/tweets/search/recent?query=from:" + twitterHandle + "&max_results=" + maxResults}
          headers:
            Authorization: ${"Bearer " + bearerToken}
        result: searchTweetsResult

In processPosts steps, analyze each Tweet in a for-in loop using the Language API connector and keep track of the current, total and min sentiment score:

    - processPosts:
        for:
          value: tweet
          in: ${searchTweetsResult.body.data}
          steps:
            - analyzeSentiment:
                call: googleapis.language.v1.documents.analyzeSentiment
                args:
                    body:
                      document:
                        content: ${tweet.text}
                        type: "PLAIN_TEXT"
                result: sentimentResult
            - updateTotalSentimentScore:
                assign:
                    - currentScore: ${sentimentResult.documentSentiment.score}
                    - totalSentimentScore: ${totalSentimentScore + currentScore}
            - updateMinSentiment:
                switch:
                  - condition: ${currentScore < minSentimentScore}
                    steps:
                      - assignMinSentiment:
                          assign:
                            - minSentimentScore: ${currentScore}
                            - minSentimentIndex: ${tweetIndex}

In the last steps, calculate the average sentiment score and return the results:

    - assignResult:
        assign:
          - numberOfTweets: ${len(searchTweetsResult.body.data)}
          - averageSentiment: ${totalSentimentScore / numberOfTweets}
    - logResult:
        call: sys.log
        args:
          text: ${"N:" + string(numberOfTweets) + " tweets with average sentiment:" + string(averageSentiment) + " min sentiment:" + string(minSentimentScore) + " at index:" + string(minSentimentIndex)}
    - returnResult:
        return:
          numberOfTweets: ${numberOfTweets}
          totalSentimentScore: ${totalSentimentScore}
          averageSentiment: ${averageSentiment}
          minSentimentScore: ${minSentimentScore}
          minSentimentIndex: ${minSentimentIndex}

You can see the full workflow.yaml.

Deploy and execute workflow

Deploy workflow:

gcloud workflows deploy twitter-sentiment \
    --source=workflow.yaml

Execute workflow:

gcloud workflows execute twitter-sentiment \
    --data='{"bearerToken":"", "twitterHandle":"GoogleCloudTech","maxResults":"100"}'

After a minute or so, you should see the see the result:

gcloud workflows executions describe bcf52313-4ce9-4c4f-9b5e-2f461223923f twitter-sentiment

...
result: '{"averageSentiment":0.2707692307692307,"minSentimentIndex":57,"minSentimentScore":-0.2,"numberOfTweets":65,"totalSentimentScore":17.599999999999994}'
state: SUCCEEDED