Sentiment analysis of laptop brands in India

Hello readers,

This time, we look into the twitter handles of some of the laptop brands popular in India.
Brands under consideration and their twitter handles are:

Lenovo India: @Lenovo_in
Apple: @AppleIncIndia
HP: @HPIndia
Sony: @sony_india
Samsung: @samsung_in
Dell: @Dell_IN
Asus: @asusindia
Acer: @acer_india

Approach methodology:  http://www.inside-r.org/howto/mining-twitter-airline-consumer-sentiment
R packages used: twitteR and dependencies


Results table:

Laptop brand	Tweets count	Very.pos.count	Very.neg.count	Very.tot	Score (100)
Acer	27	1	0	1	100
HP	1000	201	9	210	96
Lenovo	1000	315	17	332	95
Asus	590	208	14	222	94
Dell	744	63	9	72	88
Samsung	852	128	24	152	84
Sony	655	119	28	147	81
Apple	19	0	1	1	0

Any posts/ tweets with score greater than +2 is very positive and lesser than -2 is very negative.
We can ignore twitter handles that returned less than 100 tweets.
Score= 100*very.pos.count/very.tot

As Acer and Apple handles returned very low tweets, we can ignore them safely.

HP and Lenovo share the top with scores of 96 and 95 respectively.

Chart:

Sentiment analysis of insurance firms

Hello readers,

In this segment, the responses of people on twitter handles of insurance firms have been analysed.
Firms under consideration were:

MetLife @metlife
Northwestern Mutual @NM_News
Prudential of America @PrudentialNews
New York Life @NewYorkLife
Lincoln National @lincolnfingroup
MassMutual @massmutual
John Hancock @JohnHancockUSA
State Farm  @StateFarm
Guardian Life Insurance Co. @guardianlife

The approach methodology and R packages used can be found from my other posts on this blog.
The results of analysis are as under:

Sl. No.	Insurance firms	Total tweets returned from handle	very.pos.count	very.neg.count	very.tot	score (%)
1	Guardian Life Insurance Co.	65	9	0	9	100
8	Prudential of America	34	5	0	5	100
6	New York Life	227	46	1	47	98
9	State Farm	1000	86	14	100	86
2	John Hancock	79	12	3	15	80
3	Lincoln National	56	4	1	5	80
4	MassMutual	496	11	4	15	73
7	Northwestern Mutual	368	78	60	138	57
5	MetLife	1000	52	180	232	22
Legend:
Any posts/ tweets with score greater than +2 is very positive and lesser than -2 is very negative.
We can ignore twitter handles that returned less than 100 tweets.
Score= 100*very.pos.count/very.tot

New York Life has appeared a very clear leader though its handle returned only 227 tweets.
State Farm and MetLife returned 1000 tweets, but it was surprising to see MetLife come last.

Plot:

Thanks.

Comments and feedback are welcome.

Sentiment analytics for travel companies

Hello readers,



This holiday season, see what the people are saying about the popular travel companies.
The firms under consideration are:

Kayak - @KAYAK
Orbitz - @Orbitz
Expedia - @Expedia
Hipmunk - @thehipmunk
Priceline.com - @priceline
TripIt! - @TripIt
Booking.com - @bookingcom
Hotels.com - @hotelsdotcom
SeatGuru - @SeatGuru
SeatExpert - @SeatExpert
TripAdvisor - @TripAdvisor

Check my other posts for approach and R packages used.

Here are the results:

Travel	very.pos.count	very.neg.count	very.tot	score
SeatGuru	5	0	5	100
Hotels.com	721	6	727	99
TripIt!	56	2	58	97
Orbitz	110	10	120	92
Hipmunk	10	1	11	91
Kayak	133	34	167	80
Priceline.com	83	32	115	72
Booking.com	73	43	116	63
TripAdvisor	73	43	116	63
Expedia	16	23	39	41
SeatExpert	0	0	0	NaN

SeatGuru and SeatExpert are websites that help in travel planning, so technically they qualify as travel companies. The low tweet count makes their position meaningless though.

I was surprised by the high number of positive tweets about Hotels.com.

They are managing the social presence very well.

Sentiment analysis of major retailers

Hello readers,

In this holiday shopping season, let us look how the retailers have fared in the twitter sentiment analysis scores.

Following retailers were in play:

Wal-Mart @Walmart
Kroger   @kroger          
Costco   @Costco                    
Target    @Target                    
The Home Depot   @HomeDepot  
Walgreen     @Walgreens              
CVS Caremark  @cvscaremark        
Loweís        @Lowes                                    

CVS's twitter handle returned only 1 tweet, so had to remove it from final results.

Approach used: http://www.inside-r.org/howto/mining-twitter-airline-consumer-sentiment

R package used: twitteR

Output:

	Retailer	Very.pos.count	Very.neg.count	Very.tot	Score
	Target	230	38	268	86
	Kroger	155	35	190	82
	Walmart	133	34	167	80
	Costco	69	19	88	78
	Walgreen	33	13	46	72
	The Home Depot	86	49	135	64
	Lowes	91	194	285	32

Sentiment analysis for Indian e-commerce websites as on 20/12/15

Hello readers,


The sentiment analysis algorithm reads the latest 1000 tweets posted on the official twitter handles of the following e-retailers and then analyses the consumer mood about them:
 1: Amazon (@amazonIn)
 2: eBay (@eBay)
 3: Flipkart (@Flipkart)
 4: Jabong India (@JabongIndia)
 5: Myntra (@myntra)
 6: Snapdeal (@snapdeal)

Approach used: http://www.inside-r.org/howto/mining-twitter-airline-consumer-sentiment
R packages used: twitteR and associated dependencies

Here is the result table:

Sl. No.			very.pos.count	very.neg.count	very.tot	Score
1	Amazon		81	45	126	64
2	eBay		89	13	102	87
3	Flipkart		69	13	82	84
4	Jabong India		109	9	118	92
5	Myntra		100	81	181	55
6	Snapdeal		66	83	149	44