[SEARCH] [CONTEXT] Customize your search results
- Olalla Salazar (Unlicensed)
In software development and computer systems there has always been a conflict between the little “humanity” of the engines and the interactions with it.
Social analysis brings us an undeniable statement: we are all different.
Although fashion displays a cyclic model in which seasons and styles are standardized for most people, we must always consider that the user has his/her personal tastes and that this should be reflected in their interaction with computer systems. The same should happen with search engines and their results, no search is the same as any person is, so we must try to make people feel unique and special when performing their searches so they have a feeling that the search engine “understands” them.
To achieve this, at empathy.co we have developed Context! It ’s a customisation system that will allow us to generate an user profile, so that the searches fit the client’s tastes and preferences. You can find more about Context functionalities here.
HOW TO APPLY CONTEXTUALISATION TO YOUR SEARCHES
The first step is to _activate the contextualisation for your searches. O_ur integrations experts will cset up the most interesting fields for each use case and, from that moment, interactions and searches of each user will generate a totally anonymous preference profile that will allow personalising the results.
So … tell me what you want, what you really, really want.
In the following visual example we can see how Mel C and Vicky achieve different results in the search for jeans thanks to that customisation.
SO … HOW DO I ACCESS CONTEXTUALISATION DATA?
These data, also called affinity model, can be obtained thanks to the anonymous user identifier that is generated when entering e-commerce and is divided into two parts; user preferences, related to the features of the products seen and purchased during each search, and a list of the products visited by the user. In the following script we can access this data with a small and simple script made in GO:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 | package main import ( "fmt" "net/http" "net/url" "encoding/json" ) const ( //Header contentType = "Content-Type" applicationJson = "application/json" //URLs userContextUrl = "https://api.empathybroker.com/contextsearch/v2/usercontext/all/{your_site}" userAffModelUrl = "https://api.empathybroker.com/contextsearch/v2/usercontext/preferences/{your_site}" userLastProdUrl = "https://api.empathybroker.com/contextsearch/v2/usercontext/products/{your_site}" //Request constants userParam = "user" queryParam = "query" rowsParam = "rows" ) var httpClient = &http.Client{} //Handle errors in http request and print status and headers func handleHttpResponse(resp *http.Response, err error) { //Handle error if err != nil { panic(err) } //Print details fmt.Println("Http Response Status: ", resp.Status) fmt.Println("Http Response Headers: ", resp.Header) //Print response var result map[string]interface{} json.NewDecoder(resp.Body).Decode(&result) fmt.Println("Response: ", result) } //Method to get all the user affinity model preferences func getUserAffinityModel(user string, rows string) { baseUrl, err := url.Parse(userAffModelUrl) if err != nil { fmt.Println("Malformed URL: ", err.Error()) return } //Complete query params params := url.Values{} params.Add(userParam, user) params.Add(rowsParam, rows) // Add query params to the URL baseUrl.RawQuery = params.Encode() //Do http request req, _ := http.NewRequest(http.MethodGet, baseUrl.String(), nil) req.Header.Set(contentType, applicationJson) resp, err := httpClient.Do(req) handleHttpResponse(resp, err) } //Method to get the user last products viewed func getUserLastProductsViewed(user string, query string, rows string) { baseUrl, err := url.Parse(userAffModelUrl) if err != nil { fmt.Println("Malformed URL: ", err.Error()) return } //Complete query params params := url.Values{} params.Add(userParam, user) params.Add(queryParam, query) params.Add(rowsParam, rows) // Add query params to the URL baseUrl.RawQuery = params.Encode() //Do http request req, _ := http.NewRequest(http.MethodGet, baseUrl.String(), nil) req.Header.Set(contentType, applicationJson) resp, err := httpClient.Do(req) handleHttpResponse(resp, err) } //Method to get all the user personalisation model func getUserContext(user string, query string, rows string) { baseUrl, err := url.Parse(userContextUrl) if err != nil { fmt.Println("Malformed URL: ", err.Error()) return } //Complete query params params := url.Values{} params.Add(userParam, user) params.Add(queryParam, query) params.Add(rowsParam, rows) // Add query params to the URL baseUrl.RawQuery = params.Encode() //Do http request req, _ := http.NewRequest(http.MethodGet, baseUrl.String(), nil) req.Header.Set(contentType, applicationJson) resp, err := httpClient.Do(req) handleHttpResponse(resp, err) } func main() { //Get user affinity model getUserAffinityModel("userId", "2") //Get user last products viewed getUserLastProductsViewed("userId", "query", "2") //Get complete user context getUserContext("userId", "query", "2") } |
Our customisation system also includes a system that allows you to customise without the user personal information, since we will analyze the general data of the searches and their interactions for ALL users. In this way we will generate a general affinity model as well as a list of the most viewed products for each search, so that we can obtain these preferences for each search and offer the user much more adjusted results to the trends. In the following script we can access this data with a small and simple script made in GO:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 | package main import ( "fmt" "net/http" "net/url" "encoding/json" ) const ( //Header contentType = "Content-Type" applicationJson = "application/json" //URLs queryContextUrl = "https://api.empathybroker.com/contextsearch/v2/querycontext/all/{your_site}" queryAffModelUrl = "https://api.empathybroker.com/contextsearch/v2/querycontext/attributes/{your_site}" queryLastProdUrl = "https://api.empathybroker.com/contextsearch/v2/querycontext/products/{your_site}" //Request constants queryParam = "query" rowsParam = "rows" ) var httpClient = &http.Client{} //Handle errors in http request and print status and headers func handleHttpResponse(resp *http.Response, err error) { //Handle error if err != nil { panic(err) } //Print details fmt.Println("Http Response Status: ", resp.Status) fmt.Println("Http Response Headers: ", resp.Header) //Print response var result map[string]interface{} json.NewDecoder(resp.Body).Decode(&result) fmt.Println("Response: ", result) } //Method to get all the query affinity model preferences func getQueryAffinityModel(query string, rows string) { baseUrl, err := url.Parse(queryAffModelUrl) if err != nil { fmt.Println("Malformed URL: ", err.Error()) return } //Complete query params params := url.Values{} params.Add(queryParam, query) params.Add(rowsParam, rows) // Add query params to the URL baseUrl.RawQuery = params.Encode() //Do http request req, _ := http.NewRequest(http.MethodGet, baseUrl.String(), nil) req.Header.Set(contentType, applicationJson) resp, err := httpClient.Do(req) handleHttpResponse(resp, err) } //Method to get the query top products viewed func getQueryLastProductsViewed(query string, rows string) { baseUrl, err := url.Parse(queryLastProdUrl) if err != nil { fmt.Println("Malformed URL: ", err.Error()) return } //Complete query params params := url.Values{} params.Add(queryParam, query) params.Add(rowsParam, rows) // Add query params to the URL baseUrl.RawQuery = params.Encode() //Do http request req, _ := http.NewRequest(http.MethodGet, baseUrl.String(), nil) req.Header.Set(contentType, applicationJson) resp, err := httpClient.Do(req) handleHttpResponse(resp, err) } //Method to get all the query personalisation model func getQueryContext(query string, rows string) { baseUrl, err := url.Parse(queryContextUrl) if err != nil { fmt.Println("Malformed URL: ", err.Error()) return } //Complete query params params := url.Values{} params.Add(queryParam, query) params.Add(rowsParam, rows) // Add query params to the URL baseUrl.RawQuery = params.Encode() //Do http request req, _ := http.NewRequest(http.MethodGet, baseUrl.String(), nil) req.Header.Set(contentType, applicationJson) resp, err := httpClient.Do(req) handleHttpResponse(resp, err) } func main() { //Get query affinity model getQueryAffinityModel("query", "2") //Get query last products viewed getQueryLastProductsViewed("query", "2") //Get complete query context getQueryContext("query", "2") } |
Sounds good right? This service can be consumed and integrated into your own search system, but it is also fully integrated into ours, so you can start using contextualisation now!